Tighten execution projection indexes (codex)

[numnx]

🧩 Task: Tighten execution projection indexes

Automated task PR opened by Code UX for T04.

📋 Summary

Sprint: Sprint 142 · QS: Code Quality & Performance Audit · Task: T04

Audit and improve Code UX quality and performance by tightening existing hot-path contracts across live snapshots, execution projection, provider telemetry, dashboard render derivation, repository boundaries, MCP validation, guardrails, and documentation.

📝 Full Task Prompt

## Objective
Audit and harden the SQL/index contract for project execution snapshots, focusing on bounded slice queries and indexed filters used by live dashboard payload assembly.

## Scope
- Edit `src/repositories/execution/project-execution-snapshot-query.ts`
- Edit relevant slice modules as needed: `src/repositories/execution/execution-sprint-runs-query.ts`, `src/repositories/execution/execution-task-dispatches-query.ts`, `src/repositories/execution/execution-runtime-events-query.ts`, `src/repositories/execution/execution-invocations-query.ts`, `src/repositories/execution/execution-human-intervention-query.ts`
- Edit `src/repositories/db/app-db-migrations.ts` and `src/repositories/db/app-db-schema.ts` only for missing scalar indexes discovered in the owner queries
- Edit `tests/backend/repositories/execution/project-execution-snapshot-query.test.ts` and `tests/backend/repositories/db/app-db-schema.test.ts`
- Update `docs/architecture/code-quality-performance-contracts.md` indexed projection slices section with any new invariant

## Implementation Requirements
1. Review each query used by `queryProjectExecutionSnapshot` and ensure selected-sprint, expanded-run, project-recent, dispatch, runtime-event, invocation, and active-attention reads are visibly bounded and use stable scalar filters.
2. Add any missing `ensureIndex` and base schema index definitions needed by those reads; avoid indexes over prompt text, transcripts, markdown, large JSON blobs, or provider output.
3. Add or extend tests that create older unrelated rows plus active selected-sprint/expanded-run rows and assert snapshot results remain scoped, deduplicated, and capped.
4. Where practical in tests, assert the expected index exists for new migration/schema entries rather than relying on reviewer inspection.

## Constraints
- Do not change `ExecutionDashboardSnapshot` response shape.
- Do not introduce dashboard-side joins or additional polling endpoints.
- Preserve current usage and wall-time enrichment behavior, including deduplicating task and sprint-run IDs before rollups.
- Do not lower or remove existing caps to make tests easier.

## Verification
- Run `pnpm exec vitest run tests/backend/repositories/execution/project-execution-snapshot-query.test.ts tests/backend/repositories/db/app-db-schema.test.ts`.
- Run `pnpm run lint`.
- Success means snapshot projection remains bounded, schema/index tests cover any new indexes, and unrelated historical rows do not leak into live payload slices.

✅ QA Review

⏳ QA review runs after this PR is opened. Check the task card in the dashboard for the outcome once it lands.

🤖 Model & Provider

Provider	Codex CLI
Model	gpt-5.5

⏱️ Timing

Started	2026-07-06 18:05:04 UTC
Finished	2026-07-06 18:11:19 UTC
Duration	6m 14s

📊 Token Usage

Metric	Value
Input tokens	176,833
Cached input tokens	1,599,744
Output tokens	12,689
Total tokens	1,789,266
Invocations	1
Tool calls	49
Active time	5m 22s

Included usage estimate (subscription/local login): $2.06

🌿 Branch Info

Base: feature/CODUX-142-qs-code-quality-performance-audit-6
Head: task/feature-codux-142-qs-code-quality-performance-au-t04-codex-mr9j5nwf

---

🤖 Generated by Code UX