Indexes: add index / drop index, persistence, display (ADR-0025)

Implement ADR-0025 — indexes as a DSL DDL feature.

- Grammar: `add index [as <name>] on <T> (<cols>)`, `drop index
  <name>` / `drop index on <T> (<cols>)`, plus a `--cascade`
  flag on `drop column`.
- db.rs: index operations over the engine's native index
  catalog (no metadata table). The rebuild-table primitive now
  captures and recreates indexes, so `change column` and the
  relationship operations no longer silently drop them.
- `drop column` refuses an indexed column unless `--cascade`,
  which drops the covering indexes and reports each.
- Persistence: additive `indexes:` list in `project.yaml`
  (version unchanged); round-trips through rebuild/export/import.
- Display: an `Indexes:` section in the structure view and a
  nested tables/indexes items panel (S2).

Reconciles requirements.md (C3 index portion, S2 satisfied)
and CLAUDE.md. 1038 tests passing (+31), clippy clean.

CLAUDE.md

@@ -186,8 +186,9 @@ not yet implemented:
 - **Column drops/renames/type changes** (B2 / C2 partial): the
   rebuild-table primitive (ADR-0013) is in place; the grammar
   and dispatch are pending.
-- **Indexes** (C3 partial): `add index`, `drop index`, then
-  `EXPLAIN QUERY PLAN` rendering for QA1.
+- **Indexes**: `add index` / `drop index` done (ADR-0025).
+  `EXPLAIN QUERY PLAN` rendering for QA1 still pending (needs
+  its own QA2 rendering ADR).
 - **Modify relationship** (C3a): drop+add covers the use case
   today.
 - **m:n convenience** (C4): auto-generates a junction table

docs/adr/0025-indexes.md

@@ -0,0 +1,348 @@
+# ADR-0025: Indexes
+
+## Status
+
+Accepted
+
+## Context
+
+The requirements checklist (`C3`) commits to indexes as part
+of the schema-constraint surface, and `S2` commits to the
+items list showing "tables and per-table indexes". Neither is
+implemented yet.
+
+Indexes are the natural next teaching topic after relationships:
+they are the structure that makes `EXPLAIN QUERY PLAN` (`QA1`)
+pedagogically interesting — the plan for a filtered query
+visibly changes from a full scan to an index search once an
+index exists. `QA1` itself is a deliberate follow-up (it needs
+its own rendering ADR and a query worth explaining); this ADR
+sets it up by giving the playground real indexes.
+
+Three design problems shape the decision:
+
+1. **SQLite owns the index namespace.** Unlike foreign keys —
+   which have no name slot, the problem ADR-0013 solved with an
+   internal metadata table — an index in SQLite *is* a named
+   object. `sqlite_master` and `PRAGMA index_list` /
+   `index_info` carry the name, table, column list, and
+   uniqueness natively. There is nothing app-specific to store.
+2. **`DROP TABLE` silently drops a table's indexes.** The
+   rebuild-table primitive (ADR-0013) — used by change-column-
+   type and every relationship operation — drops and recreates
+   the table. Once indexes exist, every such operation would
+   erase them unless the primitive is taught to preserve them.
+3. **`playground.db` is a derived artifact** (ADR-0004 /
+   ADR-0015). Indexes must round-trip through `project.yaml`
+   or they vanish on `rebuild`, `export`, and `import`.
+
+## Decision
+
+### Grammar
+
+Indexes are declared and removed via DSL commands following
+ADR-0009 (required clauses keyword-based; optional names
+introduced by `as` per the ADR-0013 convention; `--` flags for
+opt-ins):
+
+```
+add index [as <name>] on <Table> (<col>[, <col>...])
+
+drop index <name>
+drop index on <Table> (<col>[, <col>...])
+```
+
+- `add index` is a third branch of the existing `add`
+  command, alongside `add column` and `add 1:n relationship`;
+  `drop index` is a new branch of the existing `drop` command.
+- `as <name>` is optional. The `as` keyword introduces the
+  name, matching `add 1:n relationship [as <name>]` (ADR-0013
+  established `as` as the convention for optional names).
+- `on <Table>` uses the keyword `on` — the SQL-natural word
+  for `CREATE INDEX ... ON table`, and pedagogically aligned.
+- The column list is parenthesised and comma-separated, the
+  same shape as `create table` and `insert`. One or more
+  columns; multiple columns produce a composite index in the
+  given order. An empty list `()` is a parse error.
+- Column-list completion resolves against the named table,
+  reusing the dynamic-subgrammar mechanism that already drives
+  `insert into T (...)` column candidates.
+
+`add unique index` is **not** part of this ADR — see
+*Out of scope*.
+
+### Auto-name format
+
+When `as <name>` is omitted, the executor generates
+`<Table>_<col1>[_<col2>...]_idx`, mirroring the descriptive,
+subject-first style of ADR-0013's relationship auto-names.
+
+Examples:
+
+- `add index on Customers (email)` → `Customers_email_idx`
+- `add index on Orders (CustId, Date)` → `Orders_CustId_Date_idx`
+
+If the generated name is already taken — which happens exactly
+when the same columns of the same table are already indexed —
+the command is refused with a friendly error naming the
+existing index (a second index on an identical column set is
+redundant). A duplicate *explicit* name is likewise a friendly
+error.
+
+### Drop forms
+
+`drop index` accepts two forms, mirroring `drop relationship`:
+
+- `drop index <name>` — for users who named the index or know
+  the generated name.
+- `drop index on <Table> (<col>...)` — the positional form,
+  resolved by matching the table and exact column set against
+  the table's indexes. No match is a friendly error; more than
+  one match is an ambiguity error listing the candidates and
+  advising the user to drop by name.
+
+### Storage — no metadata table
+
+Indexes do **not** get a `__rdbms_playground_indexes` table.
+SQLite stores everything the application needs natively:
+
+- `PRAGMA index_list(<table>)` — index name, uniqueness, and
+  `origin` (`c` = `CREATE INDEX`, `u` = UNIQUE constraint,
+  `pk` = primary key).
+- `PRAGMA index_info(<index>)` — the ordered column list.
+
+The application reads indexes through these pragmas. Only
+`origin = 'c'` indexes are treated as user indexes; the
+automatic indexes SQLite creates to back primary keys and
+UNIQUE constraints are not surfaced as user indexes.
+
+This is a deliberate divergence from the ADR-0013 relationship
+precedent. Relationships needed a metadata table because SQL
+foreign keys have no name slot; indexes have one, so the
+divergence is justified — adding a metadata table would
+duplicate state SQLite already owns and create a consistency
+hazard.
+
+The in-memory representation is a small structural value
+(`name`, `table`, ordered `columns`) carried by `db.rs`,
+`persistence`, and the renderer.
+
+### `project.yaml` persistence
+
+A top-level `indexes:` list is added to `project.yaml`,
+mirroring `relationships:`. Each entry records the index name,
+its table, and its ordered column list:
+
+```yaml
+indexes:
+  - name: Customers_email_idx
+    table: Customers
+    columns: [email]
+```
+
+- `version:` stays `1`. The field is additive and optional:
+  the `serde_yml` reader marks it `#[serde(default)]`, so
+  project files written before this change parse unchanged. No
+  migrator is required (the ADR-0015 §F3 framework stays
+  empty).
+- The hand-rolled writer emits `indexes: []` when there are
+  none, consistent with how `tables`/`relationships` render.
+- `SchemaSnapshot` gains an `indexes` vector alongside
+  `tables` and `relationships`.
+- `rebuild_from_text` recreates each index (via
+  `CREATE INDEX`) after the tables are built. `export` /
+  `import` carry indexes because they operate on the text
+  artifacts.
+
+### Rebuild-table interaction
+
+The `rebuild_table` primitive (ADR-0013) is extended so it no
+longer loses indexes:
+
+1. **Before** the `DROP TABLE`, capture the table's user
+   indexes structurally (name + ordered columns) via
+   `PRAGMA index_list` / `index_info`, filtered to
+   `origin = 'c'`.
+2. **After** the `ALTER TABLE ... RENAME`, recreate them with
+   `CREATE INDEX`.
+
+Recreation is parameterised by an optional column-rename map
+and a set of dropped columns, so the same primitive serves
+every caller:
+
+- **add / drop relationship**, **change column type** — the
+  column set is unchanged; indexes are recreated verbatim.
+- **rename column** — an index referencing the old column name
+  is regenerated with the new name; the index keeps its own
+  name. No error.
+- **drop column** — see below.
+
+Because indexes are captured structurally (not as raw SQL
+text), regeneration after a rename is a clean substitution
+rather than SQL string-munging.
+
+### Drop / rename column interaction
+
+- **rename column** and **change column type** preserve any
+  covering index transparently, per the rebuild rules above.
+- **drop column** is refused by default when an index covers
+  the dropped column. The error names the offending index(es)
+  and advises dropping them first. This matches the existing
+  conservative posture of `drop column`, which already refuses
+  primary-key and FK-involved columns.
+- A new `--cascade` flag on `drop column` opts in to the
+  cascading behaviour: covering indexes are dropped
+  automatically and each is reported in the result note.
+
+```
+drop column <col> from table <Table> [--cascade]
+```
+
+`Command::DropColumn` gains a `cascade: bool`. `--cascade` is
+the first destructive cascade flag in the DSL; future
+cascading drops should follow the same opt-in `--` pattern.
+Indexes that do *not* cover the dropped column are recreated
+normally regardless of the flag.
+
+### Display — structure view
+
+`render_structure` (the table-structure view in the output
+panel) gains an `Indexes:` section, rendered after the
+relationship sections and only when the table has at least one
+user index:
+
+```
+Customers
+  Id    [serial PK]
+  Email [text]
+  Indexes:
+    Customers_email_idx (Email)
+    cust_lookup (Email, Name)
+```
+
+`add index` and `drop index` return the affected table's
+description, so the auto-show pattern (ADR-0014) displays the
+updated structure — including this section — after the
+command, the same as `add column` and `add relationship`.
+
+### Display — items list (S2)
+
+The items list (left panel) becomes a nested list: each table,
+with its indexes indented beneath it.
+
+```
+Tables
+  Customers
+    Customers_email_idx
+    cust_lookup
+  Orders
+    Orders_date_idx
+```
+
+This satisfies `S2` ("the items list shows tables and
+per-table indexes; designed to extend to additional element
+kinds … without restructuring") — the nested model *is* that
+extensible structure; future kinds (relationships, views) slot
+in as further child rows.
+
+The panel's data model changes from a flat `Vec<String>` of
+table names to a structured list (table name plus its index
+names), populated by a schema refresh that now also reads
+indexes. Index rows are display-only: the current-table
+highlight behaviour is unchanged, and selecting an index row
+carries no new action in this ADR.
+
+### Errors and edge cases
+
+All user-facing strings obey the ADR-0002 rule — "the
+database" / "the engine", never the engine product name.
+
+- `add index` on a non-existent table → friendly error.
+- `add index` naming a column the table does not have →
+  friendly error naming the column.
+- Duplicate explicit index name, or an auto-name collision
+  (same table + column set) → friendly error naming the
+  existing index.
+- `drop index <name>` for an unknown name → friendly error.
+- `drop index on T(cols)` with no match → friendly error;
+  with multiple matches → ambiguity error listing candidates.
+- Internal `__rdbms_*` tables are not user tables, so the
+  table identifier never resolves to one.
+- `add index` / `drop index` are DSL DDL commands, available
+  in simple mode, appended to `history.log`, and replayable —
+  consistent with `add column` / `add relationship`.
+
+### Out of scope
+
+Explicitly excluded from this ADR:
+
+- **UNIQUE indexes** (`add unique index`). A unique index is
+  also a constraint; UNIQUE is tracked as its own `C3`
+  sub-item and is a distinct teaching concern.
+- **Partial indexes** (`CREATE INDEX ... WHERE`), **expression
+  / computed indexes**, and per-column **`DESC` / collation**
+  modifiers — advanced features beyond the playground's
+  pedagogical aim. Plain column-list indexes only.
+- **`EXPLAIN QUERY PLAN` / `QA1`** — the deliberate follow-up.
+  It needs its own rendering ADR (`QA2`) and builds on the
+  indexes this ADR delivers.
+
+## Consequences
+
+- The playground gains real, persistent indexes, advancing the
+  index portion of `C3` and satisfying `S2`.
+- The rebuild-table primitive now preserves indexes. This also
+  closes a latent bug: once indexes exist, column rename /
+  type-change would otherwise silently drop them — there are no
+  indexes today, so the bug is latent rather than live, but the
+  fix ships with the feature that would trigger it.
+- A new structural index representation threads through
+  `db.rs`, `persistence`, and `output_render`.
+- No new internal table — a deliberate divergence from the
+  ADR-0013 relationship precedent, justified by SQLite owning
+  the index namespace natively.
+- The items panel is no longer a flat list; the nested model
+  is the `S2`-mandated extension point for future element
+  kinds.
+- `drop column --cascade` establishes the opt-in `--` flag
+  pattern for destructive cascades.
+- `EXPLAIN QUERY PLAN` (`QA1`) becomes worthwhile: once it
+  lands, `show data <T> where <col> = <val>` is a query whose
+  plan visibly changes when an index on `<col>` exists.
+
+## Implementation notes
+
+Two details settled differently from the sketch above, recorded
+here so the decision text and the code agree:
+
+- **`rename column` / `drop column` do not use the rebuild
+  primitive.** Both run native `ALTER TABLE` (the playground
+  targets SQLite 3.25+/3.35+). `ALTER TABLE … RENAME COLUMN`
+  already rewrites index definitions that reference the renamed
+  column, so rename needs no index code at all. `drop column`
+  detects covering indexes directly and either refuses or, with
+  `--cascade`, issues `DROP INDEX` before the column drop. Only
+  `change column` (and the relationship operations) go through
+  the rebuild primitive, and there the column set is unchanged,
+  so the captured indexes are recreated verbatim — no
+  column-rename map or dropped-column set is needed.
+
+- **The items list keeps a flat `app.tables` plus a cache
+  map.** Rather than restructuring `app.tables` and the
+  `TablesRefreshed` event payload, per-table index names ride
+  in `SchemaCache::table_indexes`, populated by the existing
+  schema-cache refresh. The panel renders the ordered table
+  list with each table's indexes indented beneath — the
+  `S2` nested view — reading the two together.
+
+## See also
+
+- ADR-0004 / ADR-0015 (project file format and storage runtime)
+- ADR-0009 (DSL command syntax conventions)
+- ADR-0012 (internal column metadata — and why indexes diverge
+  from that precedent)
+- ADR-0013 (relationships, the rebuild-table primitive, and the
+  `as <name>` convention)
+- ADR-0014 (auto-show after writes)
+- ADR-0023 / ADR-0024 (the unified grammar tree the new
+  commands plug into)

docs/adr/README.md

@@ -30,3 +30,4 @@ This directory contains the project's ADRs, recorded per
 - [ADR-0022 — Ambient typing assistance: colour, hint panel, completion (I3 + I4)](0022-ambient-typing-assistance.md)
 - [ADR-0023 — Unified declarative grammar tree](0023-unified-grammar-tree.md) — direction (superseded for execution detail by ADR-0024)
 - [ADR-0024 — Unified grammar tree: execution plan](0024-unified-grammar-tree-execution-plan.md) — **Accepted**, the executable spec — implemented (Phases A–F; Phase F shipped "minimal", `parser.rs` retained as the router — see the ADR's Phase F implementation note)
+- [ADR-0025 — Indexes](0025-indexes.md) — **Accepted**, `add index` / `drop index`, persistence, rebuild-table preservation, and items-list display (`C3` index portion + `S2`)

docs/requirements.md

@@ -26,12 +26,12 @@ repo is pushed).
 
 ## Test baseline
 
-After ADR-0024 full implementation + the handoff-14 cleanup
-pass: **1006 passing, 0 failing, 1 ignored** (`cargo test` —
-the one ignored test is a long-standing `` ```ignore ``
-doc-test in `src/friendly/mod.rs`). Clippy clean with the
-nursery lint group enabled. (Earlier reference point, after
-B2/C2: 449 passing.)
+After ADR-0025 (indexes): **1037 passing, 0 failing, 1
+ignored** (`cargo test` — the one ignored test is a
+long-standing `` ```ignore `` doc-test in
+`src/friendly/mod.rs`). Clippy clean with the nursery lint
+group enabled. (Earlier reference points: 1006 after ADR-0024
++ the handoff-14 cleanup; 449 after B2/C2.)
 
 ---
 
@@ -47,11 +47,12 @@ B2/C2: 449 passing.)
 
 - [ ] **S1** Three-region layout: items list (left), output
   panel (right), input field (bottom).
-- [ ] **S2** Items list shows tables and per-table indexes;
+- [x] **S2** Items list shows tables and per-table indexes;
   designed to extend to additional element kinds (relations,
   views, etc.) without restructuring.
-  *(Progress: tables are listed live from the database; indexes
-  pending alongside C3 index support.)*
+  *(ADR-0025: the items panel renders a nested list — each
+  table with its index names indented beneath it. The nested
+  model is the extension point for future element kinds.)*
 - [ ] **S3** Output panel renders a visualization of the
   currently selected item and supports multiple tabs.
 - [ ] **S4** Hint area below the input field; keyboard-toggleable
@@ -130,7 +131,9 @@ B2/C2: 449 passing.)
   FK with `ON DELETE` / `ON UPDATE` actions done (ADR-0013) —
   declared via `add 1:n relationship`; symmetric outbound +
   inbound view in the structure renderer; type compatibility
-  validated at declaration via `Type::fk_target_type()`. Index,
+  validated at declaration via `Type::fk_target_type()`.
+  Indexes done (ADR-0025) — `add index` / `drop index`,
+  rebuild-preserving, persisted in `project.yaml`.
   `NOT NULL`, `UNIQUE`, `CHECK`, `DEFAULT` still pending.)*
 - [~] **C3a** Modify relationship: `modify relationship <name>
   [on delete <action>] [on update <action>]`. Users can achieve
@@ -339,6 +342,9 @@ B2/C2: 449 passing.)
 - [ ] **QA1** `EXPLAIN QUERY PLAN` is run on demand for queries;
   output is rendered as an annotated tree highlighting full
   scans, index use, and join order.
+  *(Unblocked by ADR-0025: indexes now exist, so a plan for
+  `show data <T> where <col>=<val>` visibly changes with an
+  index. Still needs the QA2 rendering ADR.)*
 - [~] **QA2** Plan rendering specifics (tree layout, annotation
   taxonomy, colour scheme) — design and ADR pending.
 

src/app.rs

@@ -14,7 +14,7 @@ use tracing::{trace, warn};
 use crate::action::Action;
 use crate::db::{
     AddColumnResult, CascadeEffect, ChangeColumnTypeResult, DataResult, DeleteResult,
-    InsertResult, TableDescription, UpdateResult,
+    DropColumnResult, InsertResult, TableDescription, UpdateResult,
 };
 use crate::dsl::{Command, ParseError, parse_command};
 use crate::event::AppEvent;
@@ -341,6 +341,10 @@ impl App {
                 self.handle_dsl_add_column_success(&command, result);
                 Vec::new()
             }
+            AppEvent::DslDropColumnSucceeded { command, result } => {
+                self.handle_dsl_drop_column_success(&command, result);
+                Vec::new()
+            }
             AppEvent::DslFailed {
                 command,
                 error,
@@ -1146,6 +1150,26 @@ impl App {
         self.current_table = Some(result.description);
     }
 
+    fn handle_dsl_drop_column_success(
+        &mut self,
+        command: &Command,
+        result: DropColumnResult,
+    ) {
+        self.note_ok_summary(command);
+        // ADR-0025: when `--cascade` removed covering indexes,
+        // name each one so the learner sees the side effect.
+        for index in &result.dropped_indexes {
+            self.note_system(crate::t!(
+                "ok.index_dropped_with_column",
+                index = index,
+            ));
+        }
+        for line in crate::output_render::render_structure(&result.description) {
+            self.note_system(line);
+        }
+        self.current_table = Some(result.description);
+    }
+
     fn handle_dsl_change_column_success(
         &mut self,
         command: &Command,
@@ -1250,7 +1274,7 @@ impl App {
         command: &Command,
         facts: crate::friendly::FailureContext,
     ) -> crate::friendly::TranslateContext {
-        use crate::dsl::{Command as C, RelationshipSelector};
+        use crate::dsl::{Command as C, IndexSelector, RelationshipSelector};
         use crate::friendly::{Operation, TranslateContext};
         let (operation, fallback_table, fallback_column) = match command {
             C::CreateTable { name, .. } => (Operation::CreateTable, Some(name.as_str()), None),
@@ -1260,7 +1284,7 @@ impl App {
                 Some(table.as_str()),
                 Some(column.as_str()),
             ),
-            C::DropColumn { table, column } => (
+            C::DropColumn { table, column, .. } => (
                 Operation::DropColumn,
                 Some(table.as_str()),
                 Some(column.as_str()),
@@ -1292,6 +1316,13 @@ impl App {
                 ),
                 RelationshipSelector::Named { .. } => (Operation::DropRelationship, None, None),
             },
+            C::AddIndex { table, .. } => (Operation::AddIndex, Some(table.as_str()), None),
+            C::DropIndex { selector } => match selector {
+                IndexSelector::Columns { table, .. } => {
+                    (Operation::DropIndex, Some(table.as_str()), None)
+                }
+                IndexSelector::Named { .. } => (Operation::DropIndex, None, None),
+            },
             C::Insert { table, .. } => (Operation::Insert, Some(table.as_str()), None),
             C::Update { table, .. } => (Operation::Update, Some(table.as_str()), None),
             C::Delete { table, .. } => (Operation::Delete, Some(table.as_str()), None),
@@ -2023,6 +2054,7 @@ mod tests {
             }],
             outbound_relationships: Vec::new(),
             inbound_relationships: Vec::new(),
+            indexes: Vec::new(),
         }
     }
 

src/completion.rs

@@ -40,11 +40,15 @@ pub struct SchemaCache {
     pub tables: Vec<String>,
     pub columns: Vec<String>,
     pub relationships: Vec<String>,
+    pub indexes: Vec<String>,
     /// Per-table column metadata with user-facing types
     /// (ADR-0024 §Phase D). Keyed by table name; lookup is
     /// case-insensitive in `columns_for_table` so the walker
     /// can resolve `Customers` regardless of how it was typed.
     pub table_columns: std::collections::HashMap<String, Vec<TableColumn>>,
+    /// Per-table user index names (ADR-0025). Keyed by table
+    /// name; drives the nested tables/indexes items panel (S2).
+    pub table_indexes: std::collections::HashMap<String, Vec<String>>,
 }
 
 /// One column's user-facing type info, scoped to a table
@@ -65,6 +69,7 @@ impl SchemaCache {
             IdentSource::Tables => &self.tables,
             IdentSource::Columns => &self.columns,
             IdentSource::Relationships => &self.relationships,
+            IdentSource::Indexes => &self.indexes,
             IdentSource::NewName | IdentSource::Types | IdentSource::Free => &[],
         }
     }
@@ -816,16 +821,23 @@ mod tests {
     }
 
     #[test]
-    fn multi_candidate_position_offers_column_and_one_to_n() {
+    fn multi_candidate_position_offers_add_subcommands() {
         // After `add ` the parser expects `column` (for
-        // `add column ...`) and `1` (the opener for
+        // `add column ...`), `index` (for `add index ...`,
+        // ADR-0025), and `1` (the opener for
         // `add 1:n relationship ...`). The completion engine
-        // surfaces both: `column` straight from the keyword
-        // expected-set, and `1:n` as a composite literal
-        // candidate so the user can Tab through to the
-        // relationship form without knowing the surface syntax.
+        // sections keyword candidates (`column`, `index`)
+        // ahead of the `1:n` composite literal, so the literal
+        // sorts last even though `add 1:n` is declared second.
         let cs = cands("add ", 4);
-        assert_eq!(cs, vec!["column".to_string(), "1:n".to_string()]);
+        assert_eq!(
+            cs,
+            vec![
+                "column".to_string(),
+                "index".to_string(),
+                "1:n".to_string(),
+            ]
+        );
     }
 
     #[test]
@@ -1039,7 +1051,10 @@ mod tests {
     }
 
     #[test]
-    fn drop_offers_three_alternatives_alphabetised() {
+    fn drop_offers_all_four_subcommands() {
+        // `drop` branches: column / relationship / table / index
+        // (ADR-0025). Candidates follow grammar declaration
+        // order, so `index` — added last — appears last.
         let cs = cands("drop ", 5);
         assert_eq!(
             cs,
@@ -1047,6 +1062,7 @@ mod tests {
                 "column".to_string(),
                 "relationship".to_string(),
                 "table".to_string(),
+                "index".to_string(),
             ],
         );
     }
@@ -1593,13 +1609,16 @@ mod tests {
             c.sort_by(|a, b| a.text.cmp(&b.text));
             c
         }
-        // `add ` exposes `column` and `1:n` — alphabetic ranker
-        // flips them.
+        // `add ` exposes `column`, `1:n` and `index` — the
+        // alphabetic ranker reorders them.
         let cache = SchemaCache::default();
         let comp = candidates_at_cursor_with("add ", 4, &cache, alphabetic_ranker)
             .expect("some completion");
         let texts: Vec<String> = comp.candidates.into_iter().map(|c| c.text).collect();
-        assert_eq!(texts, vec!["1:n".to_string(), "column".to_string()]);
+        assert_eq!(
+            texts,
+            vec!["1:n".to_string(), "column".to_string(), "index".to_string()]
+        );
     }
 
     #[test]

src/db.rs

@@ -31,7 +31,7 @@ use tokio::sync::{mpsc, oneshot};
 use tracing::{debug, info, warn};
 
 use crate::dsl::action::ReferentialAction;
-use crate::dsl::command::{ChangeColumnMode, RelationshipSelector, RowFilter};
+use crate::dsl::command::{ChangeColumnMode, IndexSelector, RelationshipSelector, RowFilter};
 use crate::dsl::ColumnSpec;
 use crate::dsl::shortid;
 use crate::dsl::types::Type;
@@ -39,8 +39,8 @@ use crate::dsl::value::{Bound, Value, ValueError};
 use crate::output_render::{Alignment, render_diagnostic_table};
 use crate::type_change;
 use crate::persistence::{
-    CellValue, ColumnSchema, Persistence, PersistenceError, RelationshipSchema, SchemaSnapshot,
-    TableSchema, TableSnapshot, decode_cell, parse_csv, parse_schema,
+    CellValue, ColumnSchema, IndexSchema, Persistence, PersistenceError, RelationshipSchema,
+    SchemaSnapshot, TableSchema, TableSnapshot, decode_cell, parse_csv, parse_schema,
 };
 use crate::project::{DATA_DIR, PROJECT_YAML};
 
@@ -64,6 +64,24 @@ pub struct TableDescription {
     /// Relationships where *this* table is the parent (some
     /// other table's column references one of ours).
     pub inbound_relationships: Vec<RelationshipEnd>,
+    /// User-created indexes on this table (ADR-0025).
+    pub indexes: Vec<IndexInfo>,
+}
+
+/// One user-created index on a table (ADR-0025).
+///
+/// Read live from the engine's native catalog
+/// (`pragma_index_list` / `pragma_index_info`); the playground
+/// keeps no separate index metadata table. Only indexes with
+/// origin `c` (a `CREATE INDEX` statement) are surfaced — the
+/// automatic indexes backing primary keys and UNIQUE
+/// constraints are not user indexes.
+#[derive(Debug, Clone, PartialEq, Eq)]
+pub struct IndexInfo {
+    pub name: String,
+    /// Indexed columns, in index order.
+    pub columns: Vec<String>,
+    pub unique: bool,
 }
 
 /// One end of a relationship as seen from the table being
@@ -220,6 +238,18 @@ pub struct AddColumnResult {
     pub client_side_notes: Vec<String>,
 }
 
+/// Outcome of a successful `drop column …` (ADR-0025).
+///
+/// `dropped_indexes` names any index removed by `--cascade`
+/// because it covered the dropped column. Empty in the common
+/// case (no covering index, or none to cascade); the runtime
+/// renders one note line per entry.
+#[derive(Debug, Clone, PartialEq, Eq)]
+pub struct DropColumnResult {
+    pub description: TableDescription,
+    pub dropped_indexes: Vec<String>,
+}
+
 /// Outcome of a successful `change column …` (ADR-0017 §6).
 ///
 /// `description` is the post-rebuild table structure (used for
@@ -397,8 +427,9 @@ enum Request {
     DropColumn {
         table: String,
         column: String,
+        cascade: bool,
         source: Option<String>,
-        reply: oneshot::Sender<Result<TableDescription, DbError>>,
+        reply: oneshot::Sender<Result<DropColumnResult, DbError>>,
     },
     RenameColumn {
         table: String,
@@ -440,6 +471,20 @@ enum Request {
         source: Option<String>,
         reply: oneshot::Sender<Result<Option<TableDescription>, DbError>>,
     },
+    /// Create an index on a table (ADR-0025).
+    AddIndex {
+        name: Option<String>,
+        table: String,
+        columns: Vec<String>,
+        source: Option<String>,
+        reply: oneshot::Sender<Result<TableDescription, DbError>>,
+    },
+    /// Drop an index by name or by table + column set (ADR-0025).
+    DropIndex {
+        selector: IndexSelector,
+        source: Option<String>,
+        reply: oneshot::Sender<Result<TableDescription, DbError>>,
+    },
     Insert {
         table: String,
         columns: Option<Vec<String>>,
@@ -607,12 +652,48 @@ impl Database {
         &self,
         table: String,
         column: String,
+        cascade: bool,
         source: Option<String>,
-    ) -> Result<TableDescription, DbError> {
+    ) -> Result<DropColumnResult, DbError> {
         let (reply, recv) = oneshot::channel();
         self.send(Request::DropColumn {
             table,
             column,
+            cascade,
+            source,
+            reply,
+        })
+        .await?;
+        recv.await.map_err(|_| DbError::WorkerGone)?
+    }
+
+    pub async fn add_index(
+        &self,
+        name: Option<String>,
+        table: String,
+        columns: Vec<String>,
+        source: Option<String>,
+    ) -> Result<TableDescription, DbError> {
+        let (reply, recv) = oneshot::channel();
+        self.send(Request::AddIndex {
+            name,
+            table,
+            columns,
+            source,
+            reply,
+        })
+        .await?;
+        recv.await.map_err(|_| DbError::WorkerGone)?
+    }
+
+    pub async fn drop_index(
+        &self,
+        selector: IndexSelector,
+        source: Option<String>,
+    ) -> Result<TableDescription, DbError> {
+        let (reply, recv) = oneshot::channel();
+        self.send(Request::DropIndex {
+            selector,
             source,
             reply,
         })
@@ -1021,6 +1102,7 @@ fn handle_request(conn: &Connection, persistence: Option<&Persistence>, req: Req
         Request::DropColumn {
             table,
             column,
+            cascade,
             source,
             reply,
         } => {
@@ -1030,6 +1112,7 @@ fn handle_request(conn: &Connection, persistence: Option<&Persistence>, req: Req
                 source.as_deref(),
                 &table,
                 &column,
+                cascade,
             ));
         }
         Request::RenameColumn {
@@ -1119,6 +1202,34 @@ fn handle_request(conn: &Connection, persistence: Option<&Persistence>, req: Req
                 &selector,
             ));
         }
+        Request::AddIndex {
+            name,
+            table,
+            columns,
+            source,
+            reply,
+        } => {
+            let _ = reply.send(do_add_index(
+                conn,
+                persistence,
+                source.as_deref(),
+                name.as_deref(),
+                &table,
+                &columns,
+            ));
+        }
+        Request::DropIndex {
+            selector,
+            source,
+            reply,
+        } => {
+            let _ = reply.send(do_drop_index(
+                conn,
+                persistence,
+                source.as_deref(),
+                &selector,
+            ));
+        }
         Request::Insert {
             table,
             columns,
@@ -1255,6 +1366,27 @@ fn do_list_names_for(
             }
             Ok(out)
         }
+        IdentSource::Indexes => {
+            // User indexes only: a `CREATE INDEX` statement
+            // leaves a non-null `sql`, whereas the automatic
+            // indexes backing PKs / UNIQUE constraints have a
+            // null `sql`.
+            let mut stmt = conn
+                .prepare(
+                    "SELECT name FROM sqlite_master \
+                     WHERE type = 'index' AND sql IS NOT NULL \
+                     ORDER BY name;",
+                )
+                .map_err(DbError::from_rusqlite)?;
+            let rows = stmt
+                .query_map([], |row| row.get::<_, String>(0))
+                .map_err(DbError::from_rusqlite)?;
+            let mut out = Vec::new();
+            for row in rows {
+                out.push(row.map_err(DbError::from_rusqlite)?);
+            }
+            Ok(out)
+        }
         IdentSource::NewName | IdentSource::Types | IdentSource::Free => Ok(Vec::new()),
     }
 }
@@ -1426,11 +1558,22 @@ fn read_schema_snapshot(conn: &Connection) -> Result<SchemaSnapshot, DbError> {
     }
 
     let relationships = read_all_relationships(conn)?;
+    let mut indexes: Vec<IndexSchema> = Vec::new();
+    for name in &table_names {
+        for idx in read_table_indexes(conn, name)? {
+            indexes.push(IndexSchema {
+                name: idx.name,
+                table: name.clone(),
+                columns: idx.columns,
+            });
+        }
+    }
     let created_at = read_project_created_at(conn)?;
     Ok(SchemaSnapshot {
         created_at,
         tables,
         relationships,
+        indexes,
     })
 }
 
@@ -1977,7 +2120,8 @@ fn do_drop_column(
     source: Option<&str>,
     table: &str,
     column: &str,
-) -> Result<TableDescription, DbError> {
+    cascade: bool,
+) -> Result<DropColumnResult, DbError> {
     let schema = read_schema(conn, table)?;
     let col_info = schema
         .columns
@@ -2011,9 +2155,39 @@ fn do_drop_column(
         )));
     }
 
+    // Indexes covering this column (ADR-0025). Without
+    // `--cascade` a covered column is refused; with it, the
+    // covering indexes are dropped alongside the column.
+    let covering: Vec<IndexInfo> = read_table_indexes(conn, table)?
+        .into_iter()
+        .filter(|i| i.columns.iter().any(|c| c == column))
+        .collect();
+    if !covering.is_empty() && !cascade {
+        let names = covering
+            .iter()
+            .map(|i| format!("`{}`", i.name))
+            .collect::<Vec<_>>()
+            .join(", ");
+        return Err(DbError::Unsupported(format!(
+            "cannot drop `{table}.{column}` while an index covers \
+             it ({names}); drop the index first, or pass `--cascade` \
+             to drop the covering indexes too."
+        )));
+    }
+
     let tx = conn
         .unchecked_transaction()
         .map_err(DbError::from_rusqlite)?;
+    // Drop covering indexes first — the engine refuses
+    // DROP COLUMN on an indexed column otherwise. `covering`
+    // is empty unless `--cascade` was given (the refusal above).
+    for index in &covering {
+        tx.execute_batch(&format!(
+            "DROP INDEX {ident};",
+            ident = quote_ident(&index.name)
+        ))
+        .map_err(DbError::from_rusqlite)?;
+    }
     let ddl = format!(
         "ALTER TABLE {tbl} DROP COLUMN {col};",
         tbl = quote_ident(table),
@@ -2036,7 +2210,10 @@ fn do_drop_column(
     };
     finalize_persistence(conn, persistence, source, &changes)?;
     tx.commit().map_err(DbError::from_rusqlite)?;
-    Ok(description)
+    Ok(DropColumnResult {
+        description,
+        dropped_indexes: covering.into_iter().map(|i| i.name).collect(),
+    })
 }
 
 /// Rename a column.
@@ -3104,6 +3281,68 @@ fn read_schema(conn: &Connection, table: &str) -> Result<ReadSchema, DbError> {
     })
 }
 
+/// Read the user-created indexes on `table` (ADR-0025).
+///
+/// `pragma_index_list` reports every index; we keep only those
+/// with origin `c` (a `CREATE INDEX` statement) and skip partial
+/// indexes — the playground never creates partial indexes, and
+/// surfacing the automatic PK / UNIQUE indexes as user indexes
+/// would be misleading. Results are ordered by index name for
+/// stable rendering.
+fn read_table_indexes(conn: &Connection, table: &str) -> Result<Vec<IndexInfo>, DbError> {
+    let mut list_stmt = conn
+        .prepare(
+            "SELECT name, \"unique\", origin, partial \
+             FROM pragma_index_list(?1) \
+             ORDER BY name;",
+        )
+        .map_err(DbError::from_rusqlite)?;
+    let metas = list_stmt
+        .query_map([table], |row| {
+            Ok((
+                row.get::<_, String>(0)?,
+                row.get::<_, i64>(1)? != 0,
+                row.get::<_, String>(2)?,
+                row.get::<_, i64>(3)? != 0,
+            ))
+        })
+        .map_err(DbError::from_rusqlite)?;
+    let mut keep: Vec<(String, bool)> = Vec::new();
+    for meta in metas {
+        let (name, unique, origin, partial) = meta.map_err(DbError::from_rusqlite)?;
+        if origin == "c" && !partial {
+            keep.push((name, unique));
+        }
+    }
+    let mut out = Vec::with_capacity(keep.len());
+    for (name, unique) in keep {
+        let columns = read_index_columns(conn, &name)?;
+        out.push(IndexInfo {
+            name,
+            columns,
+            unique,
+        });
+    }
+    Ok(out)
+}
+
+/// The indexed columns of `index`, in index order.
+fn read_index_columns(conn: &Connection, index: &str) -> Result<Vec<String>, DbError> {
+    let mut stmt = conn
+        .prepare(
+            "SELECT name FROM pragma_index_info(?1) ORDER BY seqno;",
+        )
+        .map_err(DbError::from_rusqlite)?;
+    let rows = stmt
+        .query_map([index], |row| row.get::<_, String>(0))
+        .map_err(DbError::from_rusqlite)?;
+    let mut out = Vec::new();
+    for row in rows {
+        out.push(row.map_err(DbError::from_rusqlite)?);
+    }
+    Ok(out)
+}
+
 fn parse_action_from_sqlite(s: &str) -> ReferentialAction {
     // SQLite stores the action keywords in upper-case form
     // ("CASCADE", "SET NULL", "NO ACTION", "RESTRICT").
@@ -3270,6 +3509,13 @@ where
 
         copy_data(&tx, &temp_name, table)?;
 
+        // Capture the table's user indexes before the drop —
+        // `DROP TABLE` discards them (ADR-0025). They are
+        // recreated verbatim after the rename: every caller of
+        // this primitive preserves the column set, so the index
+        // column references stay valid.
+        let captured_indexes = read_table_indexes(&tx, table)?;
+
         tx.execute_batch(&format!(
             "DROP TABLE {ident};",
             ident = quote_ident(table)
@@ -3282,6 +3528,22 @@ where
         ))
         .map_err(DbError::from_rusqlite)?;
 
+        for index in &captured_indexes {
+            let cols = index
+                .columns
+                .iter()
+                .map(|c| quote_ident(c))
+                .collect::<Vec<_>>()
+                .join(", ");
+            let unique_kw = if index.unique { "UNIQUE " } else { "" };
+            tx.execute_batch(&format!(
+                "CREATE {unique_kw}INDEX {idx} ON {tbl} ({cols});",
+                idx = quote_ident(&index.name),
+                tbl = quote_ident(table),
+            ))
+            .map_err(DbError::from_rusqlite)?;
+        }
+
         metadata_updates(&tx)?;
 
         // Verify referential integrity before committing. Any
@@ -3597,6 +3859,167 @@ fn do_drop_relationship(
     Ok(Some(do_describe_table(conn, &parent_table)?))
 }
 
+/// Create an index on `table` over `columns` (ADR-0025).
+///
+/// Refuses a redundant index on an already-indexed column set
+/// and a name collision. The index name is auto-generated as
+/// `<table>_<col…>_idx` when not supplied.
+fn do_add_index(
+    conn: &Connection,
+    persistence: Option<&Persistence>,
+    source: Option<&str>,
+    name: Option<&str>,
+    table: &str,
+    columns: &[String],
+) -> Result<TableDescription, DbError> {
+    // 1. Table must exist; gather its columns.
+    let schema = read_schema(conn, table)?;
+    // 2. Every indexed column must exist on the table.
+    for col in columns {
+        if !schema.columns.iter().any(|c| &c.name == col) {
+            return Err(DbError::Sqlite {
+                message: format!("no such column: {table}.{col}"),
+                kind: SqliteErrorKind::NoSuchColumn,
+            });
+        }
+    }
+    // 3. Refuse a redundant index over an identical column set.
+    let existing = read_table_indexes(conn, table)?;
+    if let Some(dup) = existing
+        .iter()
+        .find(|i| i.columns.as_slice() == columns)
+    {
+        return Err(DbError::Unsupported(format!(
+            "the columns ({}) of `{table}` are already indexed by `{}`.",
+            columns.join(", "),
+            dup.name,
+        )));
+    }
+    // 4. Resolve the index name (auto-generate when omitted).
+    let resolved = name.map_or_else(
+        || format!("{table}_{}_idx", columns.join("_")),
+        ToString::to_string,
+    );
+    // 5. Refuse a name collision.
+    let name_taken: i64 = conn
+        .query_row(
+            "SELECT COUNT(*) FROM sqlite_master \
+             WHERE type = 'index' AND name = ?1;",
+            [&resolved],
+            |row| row.get(0),
+        )
+        .map_err(DbError::from_rusqlite)?;
+    if name_taken > 0 {
+        return Err(DbError::Unsupported(format!(
+            "an index named `{resolved}` already exists. \
+             Pick a different name or drop the existing one first."
+        )));
+    }
+    // 6. Create the index and persist.
+    let tx = conn
+        .unchecked_transaction()
+        .map_err(DbError::from_rusqlite)?;
+    let cols_csv = columns
+        .iter()
+        .map(|c| quote_ident(c))
+        .collect::<Vec<_>>()
+        .join(", ");
+    let ddl = format!(
+        "CREATE INDEX {idx} ON {tbl} ({cols});",
+        idx = quote_ident(&resolved),
+        tbl = quote_ident(table),
+        cols = cols_csv,
+    );
+    debug!(ddl = %ddl, "add_index");
+    tx.execute_batch(&ddl).map_err(DbError::from_rusqlite)?;
+    let description = do_describe_table(conn, table)?;
+    let changes = Changes {
+        schema_dirty: true,
+        ..Changes::default()
+    };
+    finalize_persistence(conn, persistence, source, &changes)?;
+    tx.commit().map_err(DbError::from_rusqlite)?;
+    Ok(description)
+}
+
+/// Drop an index identified by name or by table + column set
+/// (ADR-0025). Returns the affected table's description.
+fn do_drop_index(
+    conn: &Connection,
+    persistence: Option<&Persistence>,
+    source: Option<&str>,
+    selector: &IndexSelector,
+) -> Result<TableDescription, DbError> {
+    let (index_name, table_name) = match selector {
+        IndexSelector::Named { name } => {
+            let lookup = conn.query_row(
+                "SELECT tbl_name FROM sqlite_master \
+                 WHERE type = 'index' AND name = ?1 AND sql IS NOT NULL;",
+                [name],
+                |row| row.get::<_, String>(0),
+            );
+            match lookup {
+                Ok(table) => (name.clone(), table),
+                Err(rusqlite::Error::QueryReturnedNoRows) => {
+                    return Err(DbError::Sqlite {
+                        message: format!("no such index: {name}"),
+                        kind: SqliteErrorKind::Other,
+                    });
+                }
+                Err(e) => return Err(DbError::from_rusqlite(e)),
+            }
+        }
+        IndexSelector::Columns { table, columns } => {
+            // Surface a missing table as such, not as "no index".
+            read_schema(conn, table)?;
+            let matches: Vec<IndexInfo> = read_table_indexes(conn, table)?
+                .into_iter()
+                .filter(|i| i.columns.as_slice() == columns.as_slice())
+                .collect();
+            match matches.as_slice() {
+                [] => {
+                    return Err(DbError::Sqlite {
+                        message: format!(
+                            "no index on {table} ({}) exists",
+                            columns.join(", ")
+                        ),
+                        kind: SqliteErrorKind::Other,
+                    });
+                }
+                [one] => (one.name.clone(), table.clone()),
+                many => {
+                    let names = many
+                        .iter()
+                        .map(|i| format!("`{}`", i.name))
+                        .collect::<Vec<_>>()
+                        .join(", ");
+                    return Err(DbError::Unsupported(format!(
+                        "more than one index on {table} ({}) matches \
+                         ({names}); drop it by name instead.",
+                        columns.join(", ")
+                    )));
+                }
+            }
+        }
+    };
+    let tx = conn
+        .unchecked_transaction()
+        .map_err(DbError::from_rusqlite)?;
+    tx.execute_batch(&format!(
+        "DROP INDEX {ident};",
+        ident = quote_ident(&index_name)
+    ))
+    .map_err(DbError::from_rusqlite)?;
+    let description = do_describe_table(conn, &table_name)?;
+    let changes = Changes {
+        schema_dirty: true,
+        ..Changes::default()
+    };
+    finalize_persistence(conn, persistence, source, &changes)?;
+    tx.commit().map_err(DbError::from_rusqlite)?;
+    Ok(description)
+}
+
 /// Read-only wrapper around `do_describe_table` that runs an
 /// auxiliary `history.log` append for user-issued
 /// `show table` commands.
@@ -3659,12 +4082,14 @@ fn do_describe_table(conn: &Connection, name: &str) -> Result<TableDescription,
 
     let outbound_relationships = read_relationships_outbound(conn, name)?;
     let inbound_relationships = read_relationships_inbound(conn, name)?;
+    let indexes = read_table_indexes(conn, name)?;
 
     Ok(TableDescription {
         name: name.to_string(),
         columns,
         outbound_relationships,
         inbound_relationships,
+        indexes,
     })
 }
 
@@ -4396,6 +4821,25 @@ fn do_rebuild_from_text(
             load_table_csv(&tx, table, &csv_path)?;
         }
 
+        // 5b. Recreate indexes (ADR-0025). Done after the data
+        //     load — the result is identical either way, and
+        //     this keeps the structural steps (tables, FKs,
+        //     data) ahead of the derived index objects.
+        for index in &snapshot.indexes {
+            let cols = index
+                .columns
+                .iter()
+                .map(|c| quote_ident(c))
+                .collect::<Vec<_>>()
+                .join(", ");
+            tx.execute_batch(&format!(
+                "CREATE INDEX {idx} ON {tbl} ({cols});",
+                idx = quote_ident(&index.name),
+                tbl = quote_ident(&index.table),
+            ))
+            .map_err(DbError::from_rusqlite)?;
+        }
+
         // 6. Verify FK consistency before committing.
         {
             let mut check = tx
@@ -5062,11 +5506,16 @@ mod tests {
         .await
         .unwrap();
 
-        let desc = db
-            .drop_column("T".to_string(), "Score".to_string(), None)
+        let result = db
+            .drop_column("T".to_string(), "Score".to_string(), false, None)
             .await
             .unwrap();
-        let names: Vec<_> = desc.columns.iter().map(|c| c.name.as_str()).collect();
+        let names: Vec<_> = result
+            .description
+            .columns
+            .iter()
+            .map(|c| c.name.as_str())
+            .collect();
         assert_eq!(names, vec!["id"]);
 
         // Row data still accessible (id was preserved); the
@@ -5081,7 +5530,7 @@ mod tests {
         let db = db();
         make_id_table(&db, "T").await;
         let err = db
-            .drop_column("T".to_string(), "id".to_string(), None)
+            .drop_column("T".to_string(), "id".to_string(), false, None)
             .await
             .unwrap_err();
         assert!(matches!(err, DbError::Unsupported(_)), "got {err:?}");
@@ -5118,7 +5567,7 @@ mod tests {
         .unwrap();
         // Try to drop the FK column on the child side.
         let err = db
-            .drop_column("Orders".to_string(), "cust_id".to_string(), None)
+            .drop_column("Orders".to_string(), "cust_id".to_string(), false, None)
             .await
             .unwrap_err();
         assert!(matches!(err, DbError::Unsupported(_)), "got {err:?}");
@@ -5130,7 +5579,7 @@ mod tests {
         let db = db();
         make_id_table(&db, "T").await;
         let err = db
-            .drop_column("T".to_string(), "Ghost".to_string(), None)
+            .drop_column("T".to_string(), "Ghost".to_string(), false, None)
             .await
             .unwrap_err();
         match err {
@@ -5139,6 +5588,304 @@ mod tests {
         }
     }
 
+    // --- indexes (ADR-0025) -----------------------------------
+
+    /// A `serial`-PK table with one extra text `Email` column —
+    /// something indexable.
+    async fn make_indexable_table(db: &Database, name: &str) {
+        make_id_table(db, name).await;
+        db.add_column(name.to_string(), "Email".to_string(), Type::Text, None)
+            .await
+            .expect("add Email column");
+    }
+
+    #[tokio::test]
+    async fn add_index_appears_in_description() {
+        let db = db();
+        make_indexable_table(&db, "Customers").await;
+        let desc = db
+            .add_index(
+                Some("idx_email".to_string()),
+                "Customers".to_string(),
+                vec!["Email".to_string()],
+                None,
+            )
+            .await
+            .expect("add index");
+        assert_eq!(desc.indexes.len(), 1);
+        assert_eq!(desc.indexes[0].name, "idx_email");
+        assert_eq!(desc.indexes[0].columns, vec!["Email".to_string()]);
+    }
+
+    #[tokio::test]
+    async fn add_index_auto_generates_name() {
+        let db = db();
+        make_indexable_table(&db, "Customers").await;
+        let desc = db
+            .add_index(None, "Customers".to_string(), vec!["Email".to_string()], None)
+            .await
+            .expect("add index");
+        assert_eq!(desc.indexes[0].name, "Customers_Email_idx");
+    }
+
+    #[tokio::test]
+    async fn add_index_composite_auto_name_joins_columns() {
+        let db = db();
+        make_id_table(&db, "Orders").await;
+        db.add_column("Orders".to_string(), "CustId".to_string(), Type::Int, None)
+            .await
+            .unwrap();
+        db.add_column("Orders".to_string(), "Day".to_string(), Type::Date, None)
+            .await
+            .unwrap();
+        let desc = db
+            .add_index(
+                None,
+                "Orders".to_string(),
+                vec!["CustId".to_string(), "Day".to_string()],
+                None,
+            )
+            .await
+            .expect("add index");
+        assert_eq!(desc.indexes[0].name, "Orders_CustId_Day_idx");
+        assert_eq!(
+            desc.indexes[0].columns,
+            vec!["CustId".to_string(), "Day".to_string()]
+        );
+    }
+
+    #[tokio::test]
+    async fn add_index_rejects_duplicate_name() {
+        let db = db();
+        make_indexable_table(&db, "Customers").await;
+        db.add_column("Customers".to_string(), "Nick".to_string(), Type::Text, None)
+            .await
+            .unwrap();
+        db.add_index(
+            Some("idx".to_string()),
+            "Customers".to_string(),
+            vec!["Email".to_string()],
+            None,
+        )
+        .await
+        .unwrap();
+        let err = db
+            .add_index(
+                Some("idx".to_string()),
+                "Customers".to_string(),
+                vec!["Nick".to_string()],
+                None,
+            )
+            .await
+            .unwrap_err();
+        assert!(matches!(err, DbError::Unsupported(_)), "got {err:?}");
+    }
+
+    #[tokio::test]
+    async fn add_index_rejects_redundant_column_set() {
+        let db = db();
+        make_indexable_table(&db, "Customers").await;
+        db.add_index(
+            Some("a".to_string()),
+            "Customers".to_string(),
+            vec!["Email".to_string()],
+            None,
+        )
+        .await
+        .unwrap();
+        let err = db
+            .add_index(
+                Some("b".to_string()),
+                "Customers".to_string(),
+                vec!["Email".to_string()],
+                None,
+            )
+            .await
+            .unwrap_err();
+        assert!(matches!(err, DbError::Unsupported(_)), "got {err:?}");
+    }
+
+    #[tokio::test]
+    async fn add_index_rejects_missing_column() {
+        let db = db();
+        make_indexable_table(&db, "Customers").await;
+        let err = db
+            .add_index(None, "Customers".to_string(), vec!["Ghost".to_string()], None)
+            .await
+            .unwrap_err();
+        assert!(
+            matches!(
+                err,
+                DbError::Sqlite {
+                    kind: SqliteErrorKind::NoSuchColumn,
+                    ..
+                }
+            ),
+            "got {err:?}"
+        );
+    }
+
+    #[tokio::test]
+    async fn add_index_rejects_missing_table() {
+        let db = db();
+        let err = db
+            .add_index(None, "Ghost".to_string(), vec!["x".to_string()], None)
+            .await
+            .unwrap_err();
+        assert!(
+            matches!(
+                err,
+                DbError::Sqlite {
+                    kind: SqliteErrorKind::NoSuchTable,
+                    ..
+                }
+            ),
+            "got {err:?}"
+        );
+    }
+
+    #[tokio::test]
+    async fn drop_index_by_name_removes_it() {
+        let db = db();
+        make_indexable_table(&db, "Customers").await;
+        db.add_index(
+            Some("idx_email".to_string()),
+            "Customers".to_string(),
+            vec!["Email".to_string()],
+            None,
+        )
+        .await
+        .unwrap();
+        let desc = db
+            .drop_index(
+                IndexSelector::Named {
+                    name: "idx_email".to_string(),
+                },
+                None,
+            )
+            .await
+            .expect("drop index");
+        assert!(desc.indexes.is_empty());
+    }
+
+    #[tokio::test]
+    async fn drop_index_by_columns_removes_it() {
+        let db = db();
+        make_indexable_table(&db, "Customers").await;
+        db.add_index(None, "Customers".to_string(), vec!["Email".to_string()], None)
+            .await
+            .unwrap();
+        let desc = db
+            .drop_index(
+                IndexSelector::Columns {
+                    table: "Customers".to_string(),
+                    columns: vec!["Email".to_string()],
+                },
+                None,
+            )
+            .await
+            .expect("drop index");
+        assert!(desc.indexes.is_empty());
+    }
+
+    #[tokio::test]
+    async fn drop_index_unknown_name_errors() {
+        let db = db();
+        make_indexable_table(&db, "Customers").await;
+        let err = db
+            .drop_index(
+                IndexSelector::Named {
+                    name: "nope".to_string(),
+                },
+                None,
+            )
+            .await
+            .unwrap_err();
+        assert!(matches!(err, DbError::Sqlite { .. }), "got {err:?}");
+    }
+
+    #[tokio::test]
+    async fn drop_column_refuses_indexed_column_without_cascade() {
+        let db = db();
+        make_indexable_table(&db, "Customers").await;
+        db.add_index(
+            Some("idx_email".to_string()),
+            "Customers".to_string(),
+            vec!["Email".to_string()],
+            None,
+        )
+        .await
+        .unwrap();
+        let err = db
+            .drop_column("Customers".to_string(), "Email".to_string(), false, None)
+            .await
+            .unwrap_err();
+        assert!(matches!(err, DbError::Unsupported(_)), "got {err:?}");
+        assert!(format!("{err}").contains("idx_email"), "got {err}");
+    }
+
+    #[tokio::test]
+    async fn drop_column_cascade_drops_covering_index() {
+        let db = db();
+        make_indexable_table(&db, "Customers").await;
+        db.add_index(
+            Some("idx_email".to_string()),
+            "Customers".to_string(),
+            vec!["Email".to_string()],
+            None,
+        )
+        .await
+        .unwrap();
+        let result = db
+            .drop_column("Customers".to_string(), "Email".to_string(), true, None)
+            .await
+            .expect("drop column --cascade");
+        assert_eq!(result.dropped_indexes, vec!["idx_email".to_string()]);
+        assert!(result.description.indexes.is_empty());
+        assert!(
+            result
+                .description
+                .columns
+                .iter()
+                .all(|c| c.name != "Email"),
+        );
+    }
+
+    #[tokio::test]
+    async fn rebuild_table_preserves_indexes() {
+        // `change column` rebuilds the table; an index on an
+        // unrelated column must survive the rebuild (ADR-0025).
+        let db = db();
+        make_indexable_table(&db, "T").await;
+        db.add_column("T".to_string(), "Score".to_string(), Type::Int, None)
+            .await
+            .unwrap();
+        db.add_index(
+            Some("idx_email".to_string()),
+            "T".to_string(),
+            vec!["Email".to_string()],
+            None,
+        )
+        .await
+        .unwrap();
+        let result = db
+            .change_column_type(
+                "T".to_string(),
+                "Score".to_string(),
+                Type::Real,
+                ChangeColumnMode::Default,
+                None,
+            )
+            .await
+            .expect("change column type");
+        assert_eq!(result.description.indexes.len(), 1);
+        assert_eq!(result.description.indexes[0].name, "idx_email");
+        assert_eq!(
+            result.description.indexes[0].columns,
+            vec!["Email".to_string()]
+        );
+    }
+
     #[tokio::test]
     async fn rename_column_updates_schema_and_metadata() {
         let db = db();

src/dsl/command.rs

@@ -46,10 +46,14 @@ pub enum Command {
     },
     /// Remove a column from a table. Refused if the column is
     /// part of the primary key or is involved in a declared
-    /// relationship — drop the relationship first.
+    /// relationship — drop the relationship first. Refused, too,
+    /// when an index covers the column, unless `cascade` is set
+    /// (the `--cascade` flag), in which case the covering
+    /// indexes are dropped alongside the column (ADR-0025).
     DropColumn {
         table: String,
         column: String,
+        cascade: bool,
     },
     /// Rename a column. SQLite handles cascading renames in
     /// FK references on other tables; the executor mirrors
@@ -96,6 +100,19 @@ pub enum Command {
     DropRelationship {
         selector: RelationshipSelector,
     },
+    /// Create an index on one or more columns of a table
+    /// (ADR-0025). `name` is optional — when `None`, the
+    /// executor auto-generates `<Table>_<col…>_idx`.
+    AddIndex {
+        name: Option<String>,
+        table: String,
+        columns: Vec<String>,
+    },
+    /// Drop an index by name, or by positional reference to its
+    /// table and exact column set (ADR-0025).
+    DropIndex {
+        selector: IndexSelector,
+    },
     /// Re-display a table's structure in the output. Doesn't
     /// change schema; useful when the user wants to look at a
     /// table they aren't currently DDL'ing on.
@@ -253,6 +270,26 @@ impl std::fmt::Display for RelationshipSelector {
     }
 }
 
+/// How a `drop index` command identifies the index to remove
+/// (ADR-0025). Both forms are accepted; the executor resolves to
+/// a single index.
+#[derive(Debug, Clone, PartialEq, Eq)]
+pub enum IndexSelector {
+    Named { name: String },
+    Columns { table: String, columns: Vec<String> },
+}
+
+impl std::fmt::Display for IndexSelector {
+    fn fmt(&self, f: &mut std::fmt::Formatter<'_>) -> std::fmt::Result {
+        match self {
+            Self::Named { name } => write!(f, "{name}"),
+            Self::Columns { table, columns } => {
+                write!(f, "on {table} ({})", columns.join(", "))
+            }
+        }
+    }
+}
+
 impl Command {
     /// Short label for log output and result rendering.
     #[must_use]
@@ -266,6 +303,8 @@ impl Command {
             Self::ChangeColumnType { .. } => "change column",
             Self::AddRelationship { .. } => "add relationship",
             Self::DropRelationship { .. } => "drop relationship",
+            Self::AddIndex { .. } => "add index",
+            Self::DropIndex { .. } => "drop index",
             Self::ShowTable { .. } => "show table",
             Self::Insert { .. } => "insert into",
             Self::Update { .. } => "update",
@@ -318,6 +357,14 @@ impl Command {
                 // is a sensible fallback for logging.
                 RelationshipSelector::Named { name } => name,
             },
+            Self::AddIndex { table, .. } => table,
+            Self::DropIndex { selector } => match selector {
+                IndexSelector::Columns { table, .. } => table,
+                // A named drop doesn't name the table until the
+                // executor resolves it; the index name is a
+                // sensible fallback for logging.
+                IndexSelector::Named { name } => name,
+            },
             // Replay isn't tied to a single table; the path is
             // the most identifying thing for log output.
             Self::Replay { path } => path,

src/dsl/grammar/ddl.rs

@@ -12,7 +12,9 @@
 //! `parent_table` vs `child_table` for the endpoints clause).
 
 use crate::dsl::action::ReferentialAction;
-use crate::dsl::command::{ChangeColumnMode, ColumnSpec, Command, RelationshipSelector};
+use crate::dsl::command::{
+    ChangeColumnMode, ColumnSpec, Command, IndexSelector, RelationshipSelector,
+};
 use crate::dsl::grammar::{
     CommandNode, HintMode, IdentSource, Node, ValidationError, Word,
     shared::{REFERENTIAL_CLAUSES, TYPE_SLOT, TYPE_VALIDATOR},
@@ -109,6 +111,40 @@ const RELATIONSHIP_NAME_NEW: Node = Node::Hinted {
     inner: &RELATIONSHIP_NAME_NEW_IDENT,
 };
 
+const INDEX_NAME_EXISTING: Node = Node::Ident {
+    source: IdentSource::Indexes,
+    role: "index_name",
+    validator: None,
+    highlight_override: None,
+    writes_table: false,
+    writes_column: false,
+    writes_user_listed_column: false,
+};
+
+const INDEX_NAME_NEW_IDENT: Node = Node::Ident {
+    source: IdentSource::NewName,
+    role: "index_name",
+    validator: None,
+    highlight_override: None,
+    writes_table: false,
+    writes_column: false,
+    writes_user_listed_column: false,
+};
+const INDEX_NAME_NEW: Node = Node::Hinted {
+    mode: NEW_NAME_HINT,
+    inner: &INDEX_NAME_NEW_IDENT,
+};
+
+// The column list shared by `add index` / `drop index`: one or
+// more existing column names, comma-separated, inside parens.
+// `COLUMN_NAME` narrows to the `on <Table>` table's columns
+// because that ident carries `writes_table: true`.
+const INDEX_COLUMN_LIST: Node = Node::Repeated {
+    inner: &COLUMN_NAME,
+    separator: Some(&Node::Punct(',')),
+    min: 1,
+};
+
 // `[to]` and `[table]` connectives.
 const TO_OPT: Node = Node::Optional(&Node::Word(Word::keyword("to")));
 const FROM_OPT: Node = Node::Optional(&Node::Word(Word::keyword("from")));
@@ -129,6 +165,11 @@ const DROP_TABLE: Node = Node::Seq(DROP_TABLE_NODES);
 // drop_column — `drop column [from] [table] <T> : <col>`
 // =================================================================
 
+// `--cascade` (ADR-0025): opt-in to dropping any index that
+// covers the column alongside the column itself. Without it, a
+// covered column is refused with a friendly error.
+const DROP_COLUMN_CASCADE_OPT: Node = Node::Optional(&Node::Flag("cascade"));
+
 const DROP_COLUMN_NODES: &[Node] = &[
     Node::Word(Word::keyword("column")),
     FROM_OPT,
@@ -136,6 +177,7 @@ const DROP_COLUMN_NODES: &[Node] = &[
     TABLE_NAME_EXISTING,
     Node::Punct(':'),
     COLUMN_NAME,
+    DROP_COLUMN_CASCADE_OPT,
 ];
 const DROP_COLUMN: Node = Node::Seq(DROP_COLUMN_NODES);
 
@@ -213,10 +255,34 @@ const DROP_RELATIONSHIP_NODES: &[Node] = &[
 const DROP_RELATIONSHIP: Node = Node::Seq(DROP_RELATIONSHIP_NODES);
 
 // =================================================================
-// drop entry — `drop (table|column|relationship) ...`
+// drop_index — `drop index (<name> | on <T> (<col>, …))`
 // =================================================================
 
-const DROP_CHOICES: &[Node] = &[DROP_COLUMN, DROP_RELATIONSHIP, DROP_TABLE];
+const DI_POSITIONAL_NODES: &[Node] = &[
+    Node::Word(Word::keyword("on")),
+    TABLE_NAME_EXISTING,
+    Node::Punct('('),
+    INDEX_COLUMN_LIST,
+    Node::Punct(')'),
+];
+const DI_POSITIONAL: Node = Node::Seq(DI_POSITIONAL_NODES);
+
+// Positional form first — it opens with the `on` keyword, so a
+// bare index name can't be mistaken for it (mirrors DR_SELECTOR).
+const DI_SELECTOR_CHOICES: &[Node] = &[DI_POSITIONAL, INDEX_NAME_EXISTING];
+const DI_SELECTOR: Node = Node::Choice(DI_SELECTOR_CHOICES);
+
+const DROP_INDEX_NODES: &[Node] = &[
+    Node::Word(Word::keyword("index")),
+    DI_SELECTOR,
+];
+const DROP_INDEX: Node = Node::Seq(DROP_INDEX_NODES);
+
+// =================================================================
+// drop entry — `drop (table|column|relationship|index) ...`
+// =================================================================
+
+const DROP_CHOICES: &[Node] = &[DROP_COLUMN, DROP_RELATIONSHIP, DROP_TABLE, DROP_INDEX];
 const DROP_SHAPE: Node = Node::Choice(DROP_CHOICES);
 
 // =================================================================
@@ -316,10 +382,31 @@ const ADD_RELATIONSHIP_NODES: &[Node] = &[
 const ADD_RELATIONSHIP: Node = Node::Seq(ADD_RELATIONSHIP_NODES);
 
 // =================================================================
-// add entry — `add (column|1:n relationship) …`
+// add_index — `add index [as <name>] on <T> (<col>, …)`
 // =================================================================
 
-const ADD_CHOICES: &[Node] = &[ADD_COLUMN, ADD_RELATIONSHIP];
+const AI_AS_NAME_NODES: &[Node] = &[
+    Node::Word(Word::keyword("as")),
+    INDEX_NAME_NEW,
+];
+const AI_AS_NAME_OPT: Node = Node::Optional(&Node::Seq(AI_AS_NAME_NODES));
+
+const ADD_INDEX_NODES: &[Node] = &[
+    Node::Word(Word::keyword("index")),
+    AI_AS_NAME_OPT,
+    Node::Word(Word::keyword("on")),
+    TABLE_NAME_EXISTING,
+    Node::Punct('('),
+    INDEX_COLUMN_LIST,
+    Node::Punct(')'),
+];
+const ADD_INDEX: Node = Node::Seq(ADD_INDEX_NODES);
+
+// =================================================================
+// add entry — `add (column|1:n relationship|index) …`
+// =================================================================
+
+const ADD_CHOICES: &[Node] = &[ADD_COLUMN, ADD_RELATIONSHIP, ADD_INDEX];
 const ADD_SHAPE: Node = Node::Choice(ADD_CHOICES);
 
 // =================================================================
@@ -402,6 +489,18 @@ fn require_ident(path: &MatchedPath, role: &'static str) -> Result<String, Valid
         })
 }
 
+/// Every ident whose role matches, in matched (left-to-right)
+/// order. Used by the column-list commands.
+fn collect_idents(path: &MatchedPath, role: &str) -> Vec<String> {
+    path.items
+        .iter()
+        .filter_map(|i| match &i.kind {
+            MatchedKind::Ident { role: r } if *r == role => Some(i.text.clone()),
+            _ => None,
+        })
+        .collect()
+}
+
 fn parse_action(words: &[&'static str]) -> ReferentialAction {
     // `set null`, `no action`, `cascade`, `restrict`.
     if words.contains(&"set") && words.contains(&"null") {
@@ -435,7 +534,32 @@ fn build_drop(path: &MatchedPath) -> Result<Command, ValidationError> {
         Some("column") => Ok(Command::DropColumn {
             table: require_ident(path, "table_name")?,
             column: require_ident(path, "column_name")?,
+            cascade: path
+                .items
+                .iter()
+                .any(|i| matches!(&i.kind, MatchedKind::Flag("cascade"))),
         }),
+        Some("index") => {
+            // Positional form has `on` as the third Word.
+            let has_on = path
+                .items
+                .iter()
+                .any(|i| matches!(&i.kind, MatchedKind::Word("on")));
+            if has_on {
+                Ok(Command::DropIndex {
+                    selector: IndexSelector::Columns {
+                        table: require_ident(path, "table_name")?,
+                        columns: collect_idents(path, "column_name"),
+                    },
+                })
+            } else {
+                Ok(Command::DropIndex {
+                    selector: IndexSelector::Named {
+                        name: require_ident(path, "index_name")?,
+                    },
+                })
+            }
+        }
         Some("relationship") => {
             // Endpoints form has `from` as the third Word.
             let has_from = path
@@ -495,6 +619,11 @@ fn build_add(path: &MatchedPath) -> Result<Command, ValidationError> {
             })
         }
         Some("1") => build_add_relationship(path),
+        Some("index") => Ok(Command::AddIndex {
+            name: ident(path, "index_name").map(str::to_string),
+            table: require_ident(path, "table_name")?,
+            columns: collect_idents(path, "column_name"),
+        }),
         _ => Err(ValidationError {
             message_key: "parse.error_wrapper",
             args: vec![("detail", "unknown add subcommand".to_string())],
@@ -638,6 +767,7 @@ pub static DROP: CommandNode = CommandNode {
         "parse.usage.drop_table",
         "parse.usage.drop_column",
         "parse.usage.drop_relationship",
+        "parse.usage.drop_index",
     ],};
 
 pub static ADD: CommandNode = CommandNode {
@@ -645,7 +775,11 @@ pub static ADD: CommandNode = CommandNode {
     shape: ADD_SHAPE,
     ast_builder: build_add,
     help_id: Some("ddl.add"),
-    usage_ids: &["parse.usage.add_column", "parse.usage.add_relationship"],};
+    usage_ids: &[
+        "parse.usage.add_column",
+        "parse.usage.add_relationship",
+        "parse.usage.add_index",
+    ],};
 
 pub static RENAME: CommandNode = CommandNode {
     entry: Word::keyword("rename"),

src/dsl/grammar/mod.rs

@@ -67,6 +67,8 @@ pub enum IdentSource {
     Columns,
     /// Existing relationship name.
     Relationships,
+    /// Existing index name.
+    Indexes,
     /// Closed set from `Type::all()` — surfaced by the walker's
     /// content validator on column-type slots; not user-listable
     /// from the schema.
@@ -82,7 +84,10 @@ impl IdentSource {
     /// entities rather than user invention or a closed set).
     #[must_use]
     pub const fn completes_from_schema(self) -> bool {
-        matches!(self, Self::Tables | Self::Columns | Self::Relationships)
+        matches!(
+            self,
+            Self::Tables | Self::Columns | Self::Relationships | Self::Indexes
+        )
     }
 
     /// Human-facing label used in parse-error wording
@@ -97,6 +102,7 @@ impl IdentSource {
             Self::Tables => "table name",
             Self::Columns => "column name",
             Self::Relationships => "relationship name",
+            Self::Indexes => "index name",
             Self::Types => "type",
         }
     }
@@ -113,6 +119,7 @@ impl IdentSource {
             "table name" => Some(Self::Tables),
             "column name" => Some(Self::Columns),
             "relationship name" => Some(Self::Relationships),
+            "index name" => Some(Self::Indexes),
             "type" => Some(Self::Types),
             _ => None,
         }

src/dsl/mod.rs

@@ -20,8 +20,8 @@ pub mod walker;
 
 pub use action::ReferentialAction;
 pub use command::{
-    AppCommand, ChangeColumnMode, ColumnSpec, Command, MessagesValue, ModeValue,
-    RelationshipSelector, RowFilter,
+    AppCommand, ChangeColumnMode, ColumnSpec, Command, IndexSelector, MessagesValue,
+    ModeValue, RelationshipSelector, RowFilter,
 };
 pub use parser::{ParseError, parse_command};
 pub use types::Type;

src/dsl/parser.rs

@@ -235,6 +235,7 @@ fn format_expectation(e: &crate::dsl::walker::outcome::Expectation) -> String {
             IdentSource::Tables => "table name".to_string(),
             IdentSource::Columns => "column name".to_string(),
             IdentSource::Relationships => "relationship name".to_string(),
+            IdentSource::Indexes => "index name".to_string(),
             IdentSource::Types => "type".to_string(),
             IdentSource::NewName | IdentSource::Free => "identifier".to_string(),
         },
@@ -316,7 +317,7 @@ mod tests {
     use super::*;
     use crate::dsl::action::ReferentialAction;
     use crate::dsl::command::{
-        ChangeColumnMode, ColumnSpec, RelationshipSelector, RowFilter,
+        ChangeColumnMode, ColumnSpec, IndexSelector, RelationshipSelector, RowFilter,
     };
     use crate::dsl::types::Type;
     use crate::dsl::value::Value;
@@ -471,6 +472,7 @@ mod tests {
             Command::DropColumn {
                 table: "Customers".to_string(),
                 column: "Email".to_string(),
+                cascade: false,
             }
         );
     }
@@ -482,6 +484,7 @@ mod tests {
             Command::DropColumn {
                 table: "Customers".to_string(),
                 column: "Email".to_string(),
+                cascade: false,
             }
         );
         assert_eq!(
@@ -489,6 +492,7 @@ mod tests {
             Command::DropColumn {
                 table: "Customers".to_string(),
                 column: "Email".to_string(),
+                cascade: false,
             }
         );
         assert_eq!(
@@ -496,6 +500,7 @@ mod tests {
             Command::DropColumn {
                 table: "Customers".to_string(),
                 column: "Email".to_string(),
+                cascade: false,
             }
         );
     }
@@ -1156,6 +1161,81 @@ mod tests {
         );
     }
 
+    // --- add index / drop index (ADR-0025) ---
+
+    #[test]
+    fn add_index_named() {
+        assert_eq!(
+            ok("add index as idx_email on Customers (Email)"),
+            Command::AddIndex {
+                name: Some("idx_email".to_string()),
+                table: "Customers".to_string(),
+                columns: vec!["Email".to_string()],
+            }
+        );
+    }
+
+    #[test]
+    fn add_index_unnamed() {
+        assert_eq!(
+            ok("add index on Customers (Email)"),
+            Command::AddIndex {
+                name: None,
+                table: "Customers".to_string(),
+                columns: vec!["Email".to_string()],
+            }
+        );
+    }
+
+    #[test]
+    fn add_index_composite_columns() {
+        assert_eq!(
+            ok("add index on Orders (CustId, Date)"),
+            Command::AddIndex {
+                name: None,
+                table: "Orders".to_string(),
+                columns: vec!["CustId".to_string(), "Date".to_string()],
+            }
+        );
+    }
+
+    #[test]
+    fn drop_index_by_name() {
+        assert_eq!(
+            ok("drop index idx_email"),
+            Command::DropIndex {
+                selector: IndexSelector::Named {
+                    name: "idx_email".to_string(),
+                },
+            }
+        );
+    }
+
+    #[test]
+    fn drop_index_by_columns() {
+        assert_eq!(
+            ok("drop index on Customers (Email)"),
+            Command::DropIndex {
+                selector: IndexSelector::Columns {
+                    table: "Customers".to_string(),
+                    columns: vec!["Email".to_string()],
+                },
+            }
+        );
+    }
+
+    #[test]
+    fn drop_column_cascade_flag() {
+        assert_eq!(
+            ok("drop column Customers: Email --cascade"),
+            Command::DropColumn {
+                table: "Customers".to_string(),
+                column: "Email".to_string(),
+                cascade: true,
+            }
+        );
+    }
+
     #[test]
     fn identifier_allows_underscores_and_digits_after_start() {
         assert_eq!(

src/dsl/walker/mod.rs

@@ -884,6 +884,7 @@ mod tests {
         let want = Command::DropColumn {
             table: "Customers".to_string(),
             column: "Email".to_string(),
+            cascade: false,
         };
         assert_eq!(parse("drop column Customers: Email").unwrap(), want);
         assert_eq!(parse("drop column from Customers: Email").unwrap(), want);

src/event.rs

@@ -9,7 +9,7 @@ use crossterm::event::KeyEvent;
 
 use crate::db::{
     AddColumnResult, ChangeColumnTypeResult, DataResult, DbError, DeleteResult,
-    InsertResult, TableDescription, UpdateResult,
+    DropColumnResult, InsertResult, TableDescription, UpdateResult,
 };
 use crate::dsl::Command;
 
@@ -56,6 +56,13 @@ pub enum AppEvent {
         command: Command,
         result: AddColumnResult,
     },
+    /// A `drop column …` succeeded. `result` carries the
+    /// post-drop description plus the names of any indexes
+    /// removed by `--cascade` (ADR-0025).
+    DslDropColumnSucceeded {
+        command: Command,
+        result: DropColumnResult,
+    },
     /// A DSL command failed. `error` is the structured
     /// payload, `facts` is the runtime-built schema-resolved
     /// enrichment (parent tables, attempted values,

src/friendly/keys.rs

@@ -198,11 +198,13 @@ pub const KEYS_AND_PLACEHOLDERS: &[(&str, &[&str])] = &[
     // code, not the catalog, because spacing is alignment-
     // sensitive in the multi-entry case.
     ("parse.usage.add_column", &[]),
+    ("parse.usage.add_index", &[]),
     ("parse.usage.add_relationship", &[]),
     ("parse.usage.change_column", &[]),
     ("parse.usage.create_table", &[]),
     ("parse.usage.delete", &[]),
     ("parse.usage.drop_column", &[]),
+    ("parse.usage.drop_index", &[]),
     ("parse.usage.drop_relationship", &[]),
     ("parse.usage.drop_table", &[]),
     ("parse.usage.insert", &[]),
@@ -394,6 +396,7 @@ pub const KEYS_AND_PLACEHOLDERS: &[(&str, &[&str])] = &[
         &["table", "column", "src_ty", "target_ty", "total"],
     ),
     // ---- DSL command success summaries (ADR-0019 §9 sweep) ----
+    ("ok.index_dropped_with_column", &["index"]),
     ("ok.rows_deleted", &["count"]),
     ("ok.rows_inserted", &["count"]),
     ("ok.rows_updated", &["count"]),

src/friendly/strings/en-US.yaml

@@ -251,13 +251,17 @@ help:
         create table <T> with pk [<col>:<type>, ...]    — create a table
     drop: |-
         drop table <T>                                  — remove a table
-        drop column [from] [table] <T>: <col>           — remove a column
+        drop column [from] [table] <T>: <col> [--cascade]   — remove a column
+          (--cascade also drops any index that covers the column)
         drop relationship <name>                        — remove a relationship
+        drop index <name>                               — remove an index
+        drop index on <T> (<col>, ...)                  — remove an index by its columns
     add: |-
         add column [to] [table] <T>: <col> (<type>)     — add a column
           (for serial/shortid on a non-empty table: existing rows auto-filled)
         add 1:n relationship [as <name>] from <P>.<col> to <C>.<col>
           [on delete <action>] [on update <action>] [--create-fk]    — declare a relationship
+        add index [as <name>] on <T> (<col>, ...)       — create an index
     rename: |-
         rename column [in] [table] <T>: <old> to <new>  — rename a column
     change: |-
@@ -412,12 +416,16 @@ parse:
     drop_relationship: |-
       drop relationship <Name>
       drop relationship from <Parent>.<col> to <Child>.<col>
+    drop_index: |-
+      drop index <Name>
+      drop index on <Table> (<col>[, ...])
     add_column: "add column [to] [table] <Table>: <Name> (<Type>)"
     add_relationship: |-
       add 1:n relationship [as <Name>]
         from <Parent>.<col> to <Child>.<col>
         [on delete <action>] [on update <action>]
         [--create-fk]
+    add_index: "add index [as <Name>] on <Table> (<col>[, ...])"
     rename_column: "rename column [in] [table] <Table>: <Old> to <New>"
     change_column: |-
       change column [in] [table] <Table>: <Name> (<Type>)
@@ -700,6 +708,9 @@ ok:
   rows_inserted: "  {count} row(s) inserted"
   rows_updated: "  {count} row(s) updated"
   rows_deleted: "  {count} row(s) deleted"
+  # Shown beneath a `drop column --cascade` summary, once per
+  # index removed because it covered the dropped column.
+  index_dropped_with_column: "  also dropped index `{index}` (it covered the column)"
 
 # ---- Client-side success notes (ADR-0017 §6, ADR-0018 §9) ------------
 client_side:

src/friendly/translate.rs

@@ -65,6 +65,8 @@ pub enum Operation {
     ChangeColumnType,
     AddRelationship,
     DropRelationship,
+    AddIndex,
+    DropIndex,
     Query,
     Rebuild,
     Replay,
@@ -92,6 +94,8 @@ impl Operation {
             Self::ChangeColumnType => "change column",
             Self::AddRelationship => "add relationship",
             Self::DropRelationship => "drop relationship",
+            Self::AddIndex => "add index",
+            Self::DropIndex => "drop index",
             Self::Query => "query",
             Self::Rebuild => "rebuild",
             Self::Replay => "replay",

src/output_render.rs

@@ -132,6 +132,19 @@ pub fn render_structure(desc: &TableDescription) -> Vec<String> {
         }
     }
 
+    // Indexes section (ADR-0025), shown only when the table
+    // carries at least one user-created index.
+    if !desc.indexes.is_empty() {
+        out.push("Indexes:".to_string());
+        for index in &desc.indexes {
+            out.push(format!(
+                "  {} ({})",
+                index.name,
+                index.columns.join(", "),
+            ));
+        }
+    }
+
     out
 }
 
@@ -331,7 +344,7 @@ fn content_row(cells: &[String], widths: &[usize], alignments: &[Alignment]) ->
 #[cfg(test)]
 mod tests {
     use super::*;
-    use crate::db::{ColumnDescription, RelationshipEnd};
+    use crate::db::{ColumnDescription, IndexInfo, RelationshipEnd};
     use crate::dsl::ReferentialAction;
     use insta::assert_snapshot;
 
@@ -548,6 +561,7 @@ mod tests {
             ],
             outbound_relationships: Vec::new(),
             inbound_relationships: Vec::new(),
+            indexes: Vec::new(),
         };
         assert_snapshot!(render_structure(&desc).join("\n"));
     }
@@ -566,6 +580,7 @@ mod tests {
                 on_delete: ReferentialAction::Cascade,
                 on_update: ReferentialAction::NoAction,
             }],
+            indexes: Vec::new(),
         };
         let out = render_structure(&desc).join("\n");
         assert!(
@@ -590,6 +605,7 @@ mod tests {
             ],
             outbound_relationships: Vec::new(),
             inbound_relationships: Vec::new(),
+            indexes: Vec::new(),
         };
         let out = render_structure(&desc).join("\n");
         // PK appears for id, NOT NULL for name, blank for nick.
@@ -597,6 +613,40 @@ mod tests {
         assert!(out.contains("│ name │ text   │ NOT NULL"), "got:\n{out}");
     }
 
+    #[test]
+    fn render_structure_shows_indexes_section() {
+        let desc = TableDescription {
+            name: "Customers".to_string(),
+            columns: vec![
+                col("id", Type::Serial, true, false),
+                col("Email", Type::Text, false, false),
+            ],
+            outbound_relationships: Vec::new(),
+            inbound_relationships: Vec::new(),
+            indexes: vec![IndexInfo {
+                name: "idx_email".to_string(),
+                columns: vec!["Email".to_string()],
+                unique: false,
+            }],
+        };
+        let out = render_structure(&desc).join("\n");
+        assert!(out.contains("Indexes:"), "got:\n{out}");
+        assert!(out.contains("idx_email (Email)"), "got:\n{out}");
+    }
+
+    #[test]
+    fn render_structure_omits_indexes_section_when_none() {
+        let desc = TableDescription {
+            name: "T".to_string(),
+            columns: vec![col("id", Type::Serial, true, false)],
+            outbound_relationships: Vec::new(),
+            inbound_relationships: Vec::new(),
+            indexes: Vec::new(),
+        };
+        let out = render_structure(&desc).join("\n");
+        assert!(!out.contains("Indexes:"), "got:\n{out}");
+    }
+
     #[test]
     fn render_structure_falls_back_to_sqlite_type_when_user_type_missing() {
         let mut desc = TableDescription {
@@ -610,6 +660,7 @@ mod tests {
             }],
             outbound_relationships: Vec::new(),
             inbound_relationships: Vec::new(),
+            indexes: Vec::new(),
         };
         let out = render_structure(&desc).join("\n");
         // The lowercase form of the SQLite type should appear.

src/persistence/mod.rs

@@ -120,6 +120,11 @@ pub struct SchemaSnapshot {
     pub created_at: String,
     pub tables: Vec<TableSchema>,
     pub relationships: Vec<RelationshipSchema>,
+    /// Indexes across all tables (ADR-0025). Carried as a flat
+    /// list mirroring `relationships`; each entry names its
+    /// table. Empty for project files written before indexes
+    /// existed — the YAML field is optional on read.
+    pub indexes: Vec<IndexSchema>,
 }
 
 #[derive(Debug, Clone, PartialEq, Eq)]
@@ -142,6 +147,15 @@ pub struct ColumnSchema {
     pub unique: bool,
 }
 
+/// One index as recorded in `project.yaml` (ADR-0025).
+#[derive(Debug, Clone, PartialEq, Eq)]
+pub struct IndexSchema {
+    pub name: String,
+    pub table: String,
+    /// The indexed columns, in index order.
+    pub columns: Vec<String>,
+}
+
 #[derive(Debug, Clone, PartialEq, Eq)]
 pub struct RelationshipSchema {
     pub name: String,
@@ -342,6 +356,7 @@ mod tests {
             created_at: "2026-05-07T14:30:12Z".to_string(),
             tables: vec![],
             relationships: vec![],
+            indexes: vec![],
         };
         p.write_schema(&schema).unwrap();
         let body = fs::read_to_string(dir.path().join(PROJECT_YAML)).unwrap();

src/persistence/yaml.rs

@@ -23,7 +23,7 @@ use serde::Deserialize;
 use crate::dsl::action::ReferentialAction;
 use crate::dsl::types::Type;
 
-use super::{ColumnSchema, RelationshipSchema, SchemaSnapshot, TableSchema};
+use super::{ColumnSchema, IndexSchema, RelationshipSchema, SchemaSnapshot, TableSchema};
 
 /// Serialize a `SchemaSnapshot` to a `project.yaml` body.
 #[must_use]
@@ -51,9 +51,31 @@ pub(super) fn serialize_schema(schema: &SchemaSnapshot) -> String {
         }
     }
 
+    if schema.indexes.is_empty() {
+        let _ = writeln!(out, "indexes: []");
+    } else {
+        let _ = writeln!(out, "indexes:");
+        for index in &schema.indexes {
+            write_index(&mut out, index);
+        }
+    }
+
     out
 }
 
+fn write_index(out: &mut String, index: &IndexSchema) {
+    let _ = writeln!(out, "  - name: {}", quote_if_needed(&index.name));
+    let _ = writeln!(out, "    table: {}", quote_if_needed(&index.table));
+    write!(out, "    columns: [").unwrap();
+    for (i, col) in index.columns.iter().enumerate() {
+        if i > 0 {
+            out.push_str(", ");
+        }
+        out.push_str(&quote_if_needed(col));
+    }
+    let _ = writeln!(out, "]");
+}
+
 fn write_table(out: &mut String, table: &TableSchema) {
     let _ = writeln!(out, "  - name: {}", quote_if_needed(&table.name));
     write!(out, "    primary_key: [").unwrap();
@@ -215,10 +237,20 @@ pub(crate) fn parse_schema(body: &str) -> Result<SchemaSnapshot, YamlError> {
             on_update,
         });
     }
+    let indexes: Vec<IndexSchema> = raw
+        .indexes
+        .into_iter()
+        .map(|i| IndexSchema {
+            name: i.name,
+            table: i.table,
+            columns: i.columns,
+        })
+        .collect();
     Ok(SchemaSnapshot {
         created_at: raw.project.created_at,
         tables,
         relationships,
+        indexes,
     })
 }
 
@@ -279,6 +311,10 @@ struct RawProject {
     tables: Vec<RawTable>,
     #[serde(default)]
     relationships: Vec<RawRelationship>,
+    /// Optional: project files written before ADR-0025 carry no
+    /// `indexes:` field and default to an empty list.
+    #[serde(default)]
+    indexes: Vec<RawIndex>,
 }
 
 #[derive(Deserialize)]
@@ -320,6 +356,13 @@ struct RawEndpoint {
     column: String,
 }
 
+#[derive(Deserialize)]
+struct RawIndex {
+    name: String,
+    table: String,
+    columns: Vec<String>,
+}
+
 #[cfg(test)]
 mod tests {
     use super::*;
@@ -355,6 +398,11 @@ mod tests {
                 on_delete: ReferentialAction::Cascade,
                 on_update: ReferentialAction::NoAction,
             }],
+            indexes: vec![IndexSchema {
+                name: "Orders_CustId_idx".to_string(),
+                table: "Orders".to_string(),
+                columns: vec!["CustId".to_string()],
+            }],
         }
     }
 
@@ -374,6 +422,9 @@ mod tests {
         assert!(body.contains("child:  { table: Orders, column: CustId }"));
         assert!(body.contains("on_delete: cascade"));
         assert!(body.contains("on_update: no_action"));
+        assert!(body.contains("- name: Orders_CustId_idx"));
+        assert!(body.contains("table: Orders"));
+        assert!(body.contains("columns: [CustId]"));
     }
 
     #[test]
@@ -382,9 +433,11 @@ mod tests {
             created_at: "2026-05-07T14:30:12Z".to_string(),
             tables: vec![],
             relationships: vec![],
+            indexes: vec![],
         });
         assert!(body.contains("tables: []"));
         assert!(body.contains("relationships: []"));
+        assert!(body.contains("indexes: []"));
     }
 
     #[test]
@@ -401,6 +454,7 @@ mod tests {
                 }],
             }],
             relationships: vec![],
+            indexes: vec![],
         });
         assert!(body.contains("- name: \"true\""));
         assert!(body.contains("{ name: \"yes\", type: bool }"));
@@ -432,6 +486,9 @@ relationships: []
         let parsed = parse_schema(body).expect("parse minimal");
         assert_eq!(parsed.tables.len(), 0);
         assert_eq!(parsed.relationships.len(), 0);
+        // A project file with no `indexes:` field (written
+        // before ADR-0025) parses with an empty index list.
+        assert_eq!(parsed.indexes.len(), 0);
         assert_eq!(parsed.created_at, "2026-05-07T14:30:12Z");
     }
 
@@ -496,6 +553,7 @@ relationships:
                 ],
             }],
             relationships: vec![],
+            indexes: vec![],
         });
         assert!(body.contains("primary_key: [a, b]"));
     }

src/runtime.rs

@@ -30,7 +30,7 @@ use crate::app::App;
 use crate::cli::Args;
 use crate::db::{
     AddColumnResult, ChangeColumnTypeResult, DataResult, Database, DbError, DeleteResult,
-    InsertResult, TableDescription, UpdateResult,
+    DropColumnResult, InsertResult, TableDescription, UpdateResult,
 };
 use crate::dsl::Command;
 use crate::event::AppEvent;
@@ -863,6 +863,9 @@ async fn build_schema_cache(database: &Database) -> crate::completion::SchemaCac
     if let Ok(rels) = database.list_names_for(IdentSource::Relationships).await {
         cache.relationships = rels;
     }
+    if let Ok(indexes) = database.list_names_for(IdentSource::Indexes).await {
+        cache.indexes = indexes;
+    }
     // Phase D (ADR-0024 §Phase D): per-table column metadata
     // with user-facing types. The walker's
     // `DynamicSubgrammar(column_value_list)` reads this to
@@ -872,6 +875,11 @@ async fn build_schema_cache(database: &Database) -> crate::completion::SchemaCac
     // walker falls back to the schemaless value-literal list.
     for name in cache.tables.clone() {
         if let Ok(desc) = database.describe_table(name.clone(), None).await {
+            // Per-table index names for the items panel (S2,
+            // ADR-0025). Captured before `desc.columns` is
+            // consumed below.
+            let index_names: Vec<String> =
+                desc.indexes.iter().map(|i| i.name.clone()).collect();
             let cols: Vec<TableColumn> = desc
                 .columns
                 .into_iter()
@@ -882,7 +890,8 @@ async fn build_schema_cache(database: &Database) -> crate::completion::SchemaCac
                     })
                 })
                 .collect();
-            cache.table_columns.insert(name, cols);
+            cache.table_columns.insert(name.clone(), cols);
+            cache.table_indexes.insert(name, index_names);
         }
     }
     cache
@@ -1039,6 +1048,10 @@ fn spawn_dsl_dispatch(
                 command: command.clone(),
                 result,
             },
+            Ok(CommandOutcome::DropColumn(result)) => AppEvent::DslDropColumnSucceeded {
+                command: command.clone(),
+                result,
+            },
             Err(DbError::PersistenceFatal {
                 operation,
                 path,
@@ -1367,6 +1380,7 @@ enum CommandOutcome {
     Delete(DeleteResult),
     ChangeColumn(ChangeColumnTypeResult),
     AddColumn(AddColumnResult),
+    DropColumn(DropColumnResult),
 }
 
 /// Spawn a task that reads a script file and dispatches each
@@ -1576,10 +1590,14 @@ async fn execute_command_typed(
             .add_column(table, column, ty, src)
             .await
             .map(CommandOutcome::AddColumn),
-        Command::DropColumn { table, column } => database
-            .drop_column(table, column, src)
+        Command::DropColumn {
+            table,
+            column,
+            cascade,
+        } => database
+            .drop_column(table, column, cascade, src)
             .await
-            .map(|d| CommandOutcome::Schema(Some(d))),
+            .map(CommandOutcome::DropColumn),
         Command::RenameColumn { table, old, new } => database
             .rename_column(table, old, new, src)
             .await
@@ -1620,6 +1638,18 @@ async fn execute_command_typed(
             .drop_relationship(selector, src)
             .await
             .map(CommandOutcome::Schema),
+        Command::AddIndex {
+            name,
+            table,
+            columns,
+        } => database
+            .add_index(name, table, columns, src)
+            .await
+            .map(|d| CommandOutcome::Schema(Some(d))),
+        Command::DropIndex { selector } => database
+            .drop_index(selector, src)
+            .await
+            .map(|d| CommandOutcome::Schema(Some(d))),
         Command::ShowTable { name } => database
             .describe_table(name, src)
             .await

src/ui.rs

@@ -420,20 +420,27 @@ fn render_items_panel(app: &App, theme: &Theme, frame: &mut Frame<'_>, area: Rec
         .as_ref()
         .map(|t| t.name.as_str())
         .unwrap_or_default();
-    let lines: Vec<Line<'_>> = app
-        .tables
-        .iter()
-        .map(|name| {
-            let style = if name == highlight {
-                Style::default()
-                    .fg(theme.fg)
-                    .add_modifier(Modifier::BOLD)
-            } else {
-                Style::default().fg(theme.fg)
-            };
-            Line::from(Span::styled(name.as_str(), style))
-        })
-        .collect();
+    // Nested tables / per-table indexes (S2, ADR-0025): each
+    // table line, with its index names indented beneath it.
+    let mut lines: Vec<Line<'_>> = Vec::new();
+    for name in &app.tables {
+        let style = if name == highlight {
+            Style::default()
+                .fg(theme.fg)
+                .add_modifier(Modifier::BOLD)
+        } else {
+            Style::default().fg(theme.fg)
+        };
+        lines.push(Line::from(Span::styled(name.as_str(), style)));
+        if let Some(indexes) = app.schema_cache.table_indexes.get(name) {
+            for index in indexes {
+                lines.push(Line::from(Span::styled(
+                    format!("  {index}"),
+                    Style::default().fg(theme.muted),
+                )));
+            }
+        }
+    }
     let paragraph = Paragraph::new(lines).block(block);
     frame.render_widget(paragraph, area);
 }
@@ -1013,6 +1020,7 @@ mod tests {
             ],
             outbound_relationships: Vec::new(),
             inbound_relationships: Vec::new(),
+            indexes: Vec::new(),
         };
         app.current_table = Some(desc);
         // Mirror what the App writes when a DSL command succeeds.
@@ -1041,4 +1049,21 @@ mod tests {
         let snapshot = render_to_string(&mut app, &theme, 80, 24);
         insta::assert_snapshot!("populated_with_table_dark", snapshot);
     }
+
+    #[test]
+    fn items_panel_nests_indexes_under_their_table() {
+        // S2 (ADR-0025): the items panel renders each table
+        // with its index names indented beneath it.
+        let mut app = App::new();
+        app.tables = vec!["Customers".to_string(), "Orders".to_string()];
+        app.schema_cache.table_indexes.insert(
+            "Customers".to_string(),
+            vec!["idx_email".to_string()],
+        );
+        let theme = Theme::dark();
+        let out = render_to_string(&mut app, &theme, 80, 24);
+        assert!(out.contains("Customers"), "table listed:\n{out}");
+        assert!(out.contains("Orders"), "table listed:\n{out}");
+        assert!(out.contains("idx_email"), "index nested in panel:\n{out}");
+    }
 }

tests/iteration3_rebuild.rs

@@ -391,3 +391,70 @@ fn rebuild_preserves_created_at_from_yaml() {
         "yaml should preserve the edited created_at:\n{final_yaml}",
     );
 }
+
+/// Indexes round-trip through `project.yaml` and a full rebuild
+/// (ADR-0025): create an index, drop the `.db`, rebuild from
+/// text, confirm the index is back.
+#[test]
+fn rebuild_restores_indexes() {
+    let data = tempdir();
+    let project_path = {
+        let project = project::open_or_create(None, Some(data.path())).unwrap();
+        let path = project.path().to_path_buf();
+        let db = Database::open_with_persistence(
+            project.db_path(),
+            Persistence::new(path.clone()),
+        )
+        .unwrap();
+        rt().block_on(async {
+            db.create_table(
+                "Customers".to_string(),
+                vec![
+                    ColumnSpec { name: "id".to_string(), ty: Type::Serial },
+                    ColumnSpec { name: "Email".to_string(), ty: Type::Text },
+                ],
+                vec!["id".to_string()],
+                Some("create table Customers with pk id:serial".to_string()),
+            )
+            .await
+            .unwrap();
+            db.add_index(
+                Some("idx_email".to_string()),
+                "Customers".to_string(),
+                vec!["Email".to_string()],
+                Some("add index as idx_email on Customers (Email)".to_string()),
+            )
+            .await
+            .unwrap();
+        });
+        drop(db);
+        drop(project);
+        path
+    };
+
+    // The index must be recorded in project.yaml — the `.db` is
+    // a derived artifact and gets discarded next.
+    let yaml = fs::read_to_string(project_path.join(project::PROJECT_YAML)).unwrap();
+    assert!(yaml.contains("idx_email"), "yaml should record the index:\n{yaml}");
+
+    fs::remove_file(project_path.join(PLAYGROUND_DB)).unwrap();
+
+    let project = project::Project::open(&project_path).unwrap();
+    let db = Database::open_with_persistence(
+        project.db_path(),
+        Persistence::new(project.path().to_path_buf()),
+    )
+    .unwrap();
+    rt().block_on(async {
+        db.rebuild_from_text(project.path().to_path_buf(), None)
+            .await
+            .expect("rebuild");
+    });
+
+    let desc = rt()
+        .block_on(async { db.describe_table("Customers".to_string(), None).await })
+        .expect("describe_table");
+    assert_eq!(desc.indexes.len(), 1, "index should survive rebuild");
+    assert_eq!(desc.indexes[0].name, "idx_email");
+    assert_eq!(desc.indexes[0].columns, vec!["Email".to_string()]);
+}

tests/typing_surface/index_ops.rs

@@ -0,0 +1,78 @@
+//! Matrix coverage for `add index [as <name>] on <T> (<col>, …)`
+//! and `drop index (<name> | on <T> (<col>, …))` (ADR-0025).
+
+use crate::typing_surface::*;
+use rdbms_playground::input_render::InputState;
+
+#[test]
+fn after_add_offers_index_branch() {
+    let schema = schema_multi_table();
+    let a = assess_at_end("add ", &schema);
+    assert!(matches!(a.state, InputState::IncompleteAtEof));
+    assert_candidate_present(&a, &["index"]);
+    crate::snap!("after_add_index_branch", a);
+}
+
+#[test]
+fn add_index_after_on_offers_table_names() {
+    let schema = schema_multi_table();
+    let a = assess_at_end("add index on ", &schema);
+    assert!(matches!(a.state, InputState::IncompleteAtEof));
+    assert_candidate_present(&a, &["Customers", "Orders"]);
+    crate::snap!("add_index_after_on", a);
+}
+
+#[test]
+fn add_index_open_paren_narrows_to_table_columns() {
+    let schema = schema_multi_table();
+    let a = assess_at_end("add index on Orders (", &schema);
+    assert!(matches!(a.state, InputState::IncompleteAtEof));
+    assert_candidate_present(&a, &["OrderId", "CustId", "Total"]);
+    assert_no_candidate_named(&a, &["id", "Name"]);
+    crate::snap!("add_index_open_paren", a);
+}
+
+#[test]
+fn complete_add_index_parses() {
+    let schema = schema_multi_table();
+    let a = assess_at_end("add index on Orders (CustId)", &schema);
+    assert!(matches!(a.state, InputState::Valid));
+    assert_eq!(a.parse_result.as_deref(), Ok("AddIndex"));
+    crate::snap!("add_index_complete", a);
+}
+
+#[test]
+fn complete_add_index_named_parses() {
+    let schema = schema_multi_table();
+    let a = assess_at_end("add index as ord_cust on Orders (CustId)", &schema);
+    assert!(matches!(a.state, InputState::Valid));
+    assert_eq!(a.parse_result.as_deref(), Ok("AddIndex"));
+    crate::snap!("add_index_named_complete", a);
+}
+
+#[test]
+fn after_drop_offers_index_branch() {
+    let schema = schema_multi_table();
+    let a = assess_at_end("drop ", &schema);
+    assert!(matches!(a.state, InputState::IncompleteAtEof));
+    assert_candidate_present(&a, &["index"]);
+    crate::snap!("drop_index_branch", a);
+}
+
+#[test]
+fn complete_drop_index_by_name_parses() {
+    let schema = schema_multi_table();
+    let a = assess_at_end("drop index some_idx", &schema);
+    assert!(matches!(a.state, InputState::Valid));
+    assert_eq!(a.parse_result.as_deref(), Ok("DropIndex"));
+    crate::snap!("drop_index_named_complete", a);
+}
+
+#[test]
+fn complete_drop_index_by_columns_parses() {
+    let schema = schema_multi_table();
+    let a = assess_at_end("drop index on Orders (CustId)", &schema);
+    assert!(matches!(a.state, InputState::Valid));
+    assert_eq!(a.parse_result.as_deref(), Ok("DropIndex"));
+    crate::snap!("drop_index_columns_complete", a);
+}

tests/typing_surface/mod.rs

@@ -32,6 +32,7 @@ pub mod create_table;
 pub mod drop_column;
 pub mod drop_relationship;
 pub mod add_relationship;
+pub mod index_ops;
 pub mod rename_change_column;
 pub mod app_commands;
 pub mod candidate_ordering;
@@ -203,6 +204,8 @@ fn command_kind_label(cmd: &rdbms_playground::dsl::Command) -> String {
         ChangeColumnType { .. } => "ChangeColumnType".into(),
         AddRelationship { .. } => "AddRelationship".into(),
         DropRelationship { .. } => "DropRelationship".into(),
+        AddIndex { .. } => "AddIndex".into(),
+        DropIndex { .. } => "DropIndex".into(),
         ShowTable { .. } => "ShowTable".into(),
         Insert { .. } => "Insert".into(),
         Update { .. } => "Update".into(),

tests/typing_surface/snapshots/typing_surface_matrix__typing_surface__add_relationship__after_add_offers_relationship_branch@after_add.snap

@@ -14,6 +14,10 @@ Assessment {
                     text: "column",
                     kind: Keyword,
                 },
+                Candidate {
+                    text: "index",
+                    kind: Keyword,
+                },
                 Candidate {
                     text: "1:n",
                     kind: Keyword,
@@ -34,6 +38,10 @@ Assessment {
                     text: "column",
                     kind: Keyword,
                 },
+                Candidate {
+                    text: "index",
+                    kind: Keyword,
+                },
                 Candidate {
                     text: "1:n",
                     kind: Keyword,

tests/typing_surface/snapshots/typing_surface_matrix__typing_surface__index_ops__add_index_after_on_offers_table_names@add_index_after_on.snap

@@ -0,0 +1,47 @@
+---
+source: tests/typing_surface/index_ops.rs
+description: "input=\"add index on \" cursor=13"
+expression: "& a"
+---
+Assessment {
+    input: "add index on ",
+    cursor: 13,
+    state: IncompleteAtEof,
+    hint: Some(
+        Candidates {
+            items: [
+                Candidate {
+                    text: "Customers",
+                    kind: Identifier,
+                },
+                Candidate {
+                    text: "Orders",
+                    kind: Identifier,
+                },
+            ],
+            selected: None,
+        },
+    ),
+    completion: Some(
+        Completion {
+            replaced_range: (
+                13,
+                13,
+            ),
+            partial_prefix: "",
+            candidates: [
+                Candidate {
+                    text: "Customers",
+                    kind: Identifier,
+                },
+                Candidate {
+                    text: "Orders",
+                    kind: Identifier,
+                },
+            ],
+        },
+    ),
+    parse_result: Err(
+        "Invalid(at_eof)",
+    ),
+}

tests/typing_surface/snapshots/typing_surface_matrix__typing_surface__index_ops__add_index_open_paren_narrows_to_table_columns@add_index_open_paren.snap

@@ -0,0 +1,55 @@
+---
+source: tests/typing_surface/index_ops.rs
+description: "input=\"add index on Orders (\" cursor=21"
+expression: "& a"
+---
+Assessment {
+    input: "add index on Orders (",
+    cursor: 21,
+    state: IncompleteAtEof,
+    hint: Some(
+        Candidates {
+            items: [
+                Candidate {
+                    text: "CustId",
+                    kind: Identifier,
+                },
+                Candidate {
+                    text: "OrderId",
+                    kind: Identifier,
+                },
+                Candidate {
+                    text: "Total",
+                    kind: Identifier,
+                },
+            ],
+            selected: None,
+        },
+    ),
+    completion: Some(
+        Completion {
+            replaced_range: (
+                21,
+                21,
+            ),
+            partial_prefix: "",
+            candidates: [
+                Candidate {
+                    text: "CustId",
+                    kind: Identifier,
+                },
+                Candidate {
+                    text: "OrderId",
+                    kind: Identifier,
+                },
+                Candidate {
+                    text: "Total",
+                    kind: Identifier,
+                },
+            ],
+        },
+    ),
+    parse_result: Err(
+        "Invalid(at_eof)",
+    ),
+}

tests/typing_surface/snapshots/typing_surface_matrix__typing_surface__index_ops__after_add_offers_index_branch@after_add_index_branch.snap

@@ -0,0 +1,55 @@
+---
+source: tests/typing_surface/index_ops.rs
+description: "input=\"add \" cursor=4"
+expression: "& a"
+---
+Assessment {
+    input: "add ",
+    cursor: 4,
+    state: IncompleteAtEof,
+    hint: Some(
+        Candidates {
+            items: [
+                Candidate {
+                    text: "column",
+                    kind: Keyword,
+                },
+                Candidate {
+                    text: "index",
+                    kind: Keyword,
+                },
+                Candidate {
+                    text: "1:n",
+                    kind: Keyword,
+                },
+            ],
+            selected: None,
+        },
+    ),
+    completion: Some(
+        Completion {
+            replaced_range: (
+                4,
+                4,
+            ),
+            partial_prefix: "",
+            candidates: [
+                Candidate {
+                    text: "column",
+                    kind: Keyword,
+                },
+                Candidate {
+                    text: "index",
+                    kind: Keyword,
+                },
+                Candidate {
+                    text: "1:n",
+                    kind: Keyword,
+                },
+            ],
+        },
+    ),
+    parse_result: Err(
+        "Invalid(at_eof)",
+    ),
+}

tests/typing_surface/snapshots/typing_surface_matrix__typing_surface__index_ops__after_drop_offers_index_branch@drop_index_branch.snap

@@ -0,0 +1,63 @@
+---
+source: tests/typing_surface/index_ops.rs
+description: "input=\"drop \" cursor=5"
+expression: "& a"
+---
+Assessment {
+    input: "drop ",
+    cursor: 5,
+    state: IncompleteAtEof,
+    hint: Some(
+        Candidates {
+            items: [
+                Candidate {
+                    text: "column",
+                    kind: Keyword,
+                },
+                Candidate {
+                    text: "relationship",
+                    kind: Keyword,
+                },
+                Candidate {
+                    text: "table",
+                    kind: Keyword,
+                },
+                Candidate {
+                    text: "index",
+                    kind: Keyword,
+                },
+            ],
+            selected: None,
+        },
+    ),
+    completion: Some(
+        Completion {
+            replaced_range: (
+                5,
+                5,
+            ),
+            partial_prefix: "",
+            candidates: [
+                Candidate {
+                    text: "column",
+                    kind: Keyword,
+                },
+                Candidate {
+                    text: "relationship",
+                    kind: Keyword,
+                },
+                Candidate {
+                    text: "table",
+                    kind: Keyword,
+                },
+                Candidate {
+                    text: "index",
+                    kind: Keyword,
+                },
+            ],
+        },
+    ),
+    parse_result: Err(
+        "Invalid(at_eof)",
+    ),
+}

tests/typing_surface/snapshots/typing_surface_matrix__typing_surface__index_ops__complete_add_index_named_parses@add_index_named_complete.snap

@@ -0,0 +1,19 @@
+---
+source: tests/typing_surface/index_ops.rs
+description: "input=\"add index as ord_cust on Orders (CustId)\" cursor=40"
+expression: "& a"
+---
+Assessment {
+    input: "add index as ord_cust on Orders (CustId)",
+    cursor: 40,
+    state: Valid,
+    hint: Some(
+        Prose(
+            "Submit with Enter",
+        ),
+    ),
+    completion: None,
+    parse_result: Ok(
+        "AddIndex",
+    ),
+}

tests/typing_surface/snapshots/typing_surface_matrix__typing_surface__index_ops__complete_add_index_parses@add_index_complete.snap

@@ -0,0 +1,19 @@
+---
+source: tests/typing_surface/index_ops.rs
+description: "input=\"add index on Orders (CustId)\" cursor=28"
+expression: "& a"
+---
+Assessment {
+    input: "add index on Orders (CustId)",
+    cursor: 28,
+    state: Valid,
+    hint: Some(
+        Prose(
+            "Submit with Enter",
+        ),
+    ),
+    completion: None,
+    parse_result: Ok(
+        "AddIndex",
+    ),
+}

tests/typing_surface/snapshots/typing_surface_matrix__typing_surface__index_ops__complete_drop_index_by_columns_parses@drop_index_columns_complete.snap

@@ -0,0 +1,19 @@
+---
+source: tests/typing_surface/index_ops.rs
+description: "input=\"drop index on Orders (CustId)\" cursor=29"
+expression: "& a"
+---
+Assessment {
+    input: "drop index on Orders (CustId)",
+    cursor: 29,
+    state: Valid,
+    hint: Some(
+        Prose(
+            "Submit with Enter",
+        ),
+    ),
+    completion: None,
+    parse_result: Ok(
+        "DropIndex",
+    ),
+}

tests/typing_surface/snapshots/typing_surface_matrix__typing_surface__index_ops__complete_drop_index_by_name_parses@drop_index_named_complete.snap

@@ -0,0 +1,19 @@
+---
+source: tests/typing_surface/index_ops.rs
+description: "input=\"drop index some_idx\" cursor=19"
+expression: "& a"
+---
+Assessment {
+    input: "drop index some_idx",
+    cursor: 19,
+    state: Valid,
+    hint: Some(
+        Prose(
+            "No such identifier: `some_idx`",
+        ),
+    ),
+    completion: None,
+    parse_result: Ok(
+        "DropIndex",
+    ),
+}

tests/walking_skeleton.rs

@@ -257,6 +257,7 @@ fn fake_table(name: &str, columns: &[(&str, Type, bool)]) -> TableDescription {
             .collect(),
         outbound_relationships: Vec::new(),
         inbound_relationships: Vec::new(),
+        indexes: Vec::new(),
     }
 }
 
@@ -422,6 +423,7 @@ fn add_relationship_flow_shows_parent_side_with_inbound_section() {
             on_delete: ReferentialAction::Cascade,
             on_update: ReferentialAction::NoAction,
         }],
+        indexes: Vec::new(),
     };
     app.update(AppEvent::DslSucceeded {
         command: Command::AddRelationship {
@@ -470,6 +472,7 @@ fn add_relationship_flow_shows_inbound_section_on_parent() {
             on_delete: ReferentialAction::Cascade,
             on_update: ReferentialAction::NoAction,
         }],
+        indexes: Vec::new(),
     };
     app.update(AppEvent::DslSucceeded {
         command: Command::AddColumn {