amduat-api/README.md

# amduat-app-next (Scaffold)

Starter project scaffold for building an app against `amduatd` v2.

## Included handoff assets

- `docs/v2-app-developer-guide.md` (compact app guide)
- `contracts/amduatd-api-contract.v2.json` (machine contract)
- `scripts/graph_client_helpers.sh` (reusable shell helpers)

## Quick start

1. Configure environment:

```sh
cp config/env.example config/env.local
```

2. Start a local daemon:

```sh
./scripts/dev_start_daemon.sh
```
`dev_start_daemon.sh` initializes the store for the selected backend and, when
`STORE_BACKEND=index`, runs a quick startup write probe. If index writes are
unhealthy, it first tries a one-time index store repair (backup + re-init) and
then falls back to `fs` if still unhealthy (configurable via
`INDEX_BACKEND_PROBE`, `INDEX_BACKEND_REPAIR`, and `INDEX_BACKEND_FALLBACK` in
`config/env.local`).

3. Run startup checks against the daemon socket:

```sh
./scripts/bootstrap_check.sh
```

4. Run sample idempotent batch ingest:

```sh
./scripts/ingest_example.sh
```

5. Run sample changes consumer (recommended):

```sh
./scripts/v2_app.sh consume-changes
```

## v2 Vertical Slice CLI

Use the integrated v2 app flow wrapper:

```sh
./scripts/v2_app.sh startup-check
./scripts/v2_app.sh ai-vertical-slice
./scripts/v2_app.sh ai-vertical-slice --skip-evals
./scripts/v2_app.sh ai-vertical-slice --auto-start-daemon
./scripts/v2_app.sh ai-agent 'doc-ai-1' 'What domain is doc-ai-1 in?' 'ms.within_domain'
./scripts/v2_app.sh ai-agent --json --require-evidence --max-steps 3 'doc-ai-1' 'What domain is doc-ai-1 in?' 'ms.within_domain'
./scripts/v2_app.sh ai-check
./scripts/v2_app.sh ai-generate 'Summarize retrieval behavior in one sentence.'
./scripts/v2_app.sh ai-generate --json 'Summarize retrieval behavior in one sentence.'
./scripts/v2_app.sh ai-answer 'doc-1' 'What topic is doc-1 in?' 'ms.within_domain'
./scripts/v2_app.sh ai-answer --json 'doc-1' 'What topic is doc-1 in?' 'ms.within_domain'
./scripts/v2_app.sh ai-answer --json --require-evidence 'doc-1' 'What topic is doc-1 in?' 'ms.within_domain'
./scripts/v2_app.sh ingest '{"idempotency_key":"k1","mode":"continue_on_error","nodes":[{"name":"doc-1"}]}'
./scripts/v2_app.sh sync-once
./scripts/v2_app.sh consume-changes --once
./scripts/v2_app.sh retrieve 'doc-1' 'ms.within_domain'
./scripts/v2_app.sh tombstone '<edge_ref>'
```

AI lane notes:

- Plan and scope guardrails: `docs/ai-plan.md`
- Deterministic seed payload: `ai/fixtures/seed_batch.json`
- Agent loop checkpoints: `ai/runs/agent-run-*.json`

Run integration coverage (requires running `amduatd` + `jq`):

```sh
./tests/integration_v2.sh
```

Run local cursor/handler semantics checks (no daemon required):

```sh
./tests/changes_consumer_410.sh
./tests/changes_consumer_handler.sh
```

Run AI connectivity and generation smoke eval (no daemon required):

```sh
./tests/ai_eval.sh
./tests/ai_answer_eval.sh
```

Run a fast end-to-end smoke (startup + ingest + sync + retrieve + tombstone):

```sh
./scripts/smoke_v2.sh
```

## Notes

- This scaffold assumes local Unix-socket access to `amduatd`.
- Graph cursors are opaque and must be persisted exactly as returned.
- `./scripts/sync_loop.sh` and `sync-loop` CLI are deprecated compatibility aliases.
- Keep `contracts/amduatd-api-contract.v2.json` in sync with upstream when you pull updates.