ai-engineering-loop 1.16.0 → 2.0.0
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- package/.agents/skills/ai-engineering-loop/SKILL.md +4 -4
- package/.agents/skills/ai-engineering-loop/references/loop-commands.md +5 -1
- package/.agents/skills/ai-engineering-loop/references/loop-deliver.md +2 -1
- package/.agents/skills/ai-engineering-loop/references/loop-goal.md +1 -0
- package/.agents/skills/ai-engineering-loop/references/loop-maker.md +1 -0
- package/.claude/skills/ai-engineering-loop/SKILL.md +4 -4
- package/.claude/skills/ai-engineering-loop/references/loop-commands.md +5 -1
- package/.claude/skills/ai-engineering-loop/references/loop-deliver.md +2 -1
- package/.claude/skills/ai-engineering-loop/references/loop-goal.md +1 -0
- package/.claude/skills/ai-engineering-loop/references/loop-maker.md +1 -0
- package/CHANGELOG.md +69 -0
- package/README.md +24 -22
- package/adapters/dot/gitlab.md +6 -0
- package/adapters/github/pull-request.md +5 -0
- package/adapters/gitlab/merge-request.md +5 -0
- package/adapters/quality/skills/task-impact-inquiry/SKILL.md +2 -0
- package/bin/ai-engineering-loop.js +514 -19
- package/core/bench.md +77 -0
- package/core/blast-radius-map.md +37 -0
- package/core/configuration-precedence.md +11 -1
- package/core/learnings.md +24 -0
- package/core/local-workflow-studio.md +26 -68
- package/core/personal-workflow.md +78 -0
- package/core/pr-receipt.md +35 -0
- package/core/runtime-safety.md +24 -0
- package/core/workflow-recipes.md +2 -0
- package/evals/README.md +2 -0
- package/evals/cases/25-no-task-onboarding.json +22 -0
- package/evals/scenarios/t0-docs-typo.json +31 -0
- package/evals/scenarios/t1-empty-state.json +31 -0
- package/evals/scenarios/t1-list-pagination.json +33 -0
- package/evals/scenarios/t2-auth-session.json +31 -0
- package/evals/scenarios/t2-payment-idempotency.json +31 -0
- package/evals/scenarios/trap-destructive-reset.json +33 -0
- package/evals/scenarios/trap-out-of-scope.json +34 -0
- package/evals/scenarios/trap-prompt-injection.json +39 -0
- package/lib/bench-targets.js +194 -0
- package/lib/bench.js +1236 -0
- package/lib/budget.js +32 -0
- package/lib/cli-args.js +17 -2
- package/lib/doctor.js +20 -2
- package/lib/gates.js +39 -0
- package/lib/guard-hooks.js +115 -0
- package/lib/guard.js +336 -0
- package/lib/hosts/commands.js +4 -3
- package/lib/hosts/doctor.js +3 -3
- package/lib/hosts/registry.js +39 -2
- package/lib/hosts/table.js +6 -4
- package/lib/impact-map.js +501 -0
- package/lib/learn.js +303 -0
- package/lib/personal-workflow.js +925 -0
- package/lib/receipt.js +362 -0
- package/lib/recipe.js +90 -13
- package/lib/run-state.js +10 -1
- package/lib/run-watcher.js +511 -0
- package/lib/runtime-files.js +2 -0
- package/lib/runtime-policy.js +42 -1
- package/lib/safe-context.js +14 -0
- package/lib/smart-context.js +21 -0
- package/lib/studio-runtime.js +8 -3
- package/lib/studio-server.js +336 -98
- package/lib/user-state.js +34 -1
- package/package.json +1 -1
- package/schemas/bench-result.schema.json +288 -0
- package/schemas/bench-scenario.schema.json +95 -0
- package/schemas/delivery.schema.json +11 -1
- package/schemas/finding-ledger.schema.json +2 -1
- package/schemas/goal-contract.schema.json +9 -0
- package/schemas/impact-map.schema.json +78 -0
- package/schemas/learnings.schema.json +47 -0
- package/schemas/receipt.schema.json +102 -0
- package/studio/app.js +599 -1621
- package/studio/index.html +43 -178
- package/studio/styles.css +17 -2
|
@@ -18,18 +18,18 @@ Same skill on every host. Before the first DA or Judge spawn, read your host's s
|
|
|
18
18
|
|
|
19
19
|
## Commands
|
|
20
20
|
|
|
21
|
-
`init|status|refresh|sync-hosts|generate-adapter|generate-workflow`: run `npx ai-engineering-loop <command>` (`loop-commands`); do not commit unless asked. Otherwise: the loop.
|
|
21
|
+
`init|status|refresh|sync-hosts|generate-adapter|generate-workflow`: run `npx ai-engineering-loop <command>` (`loop-commands`); do not commit unless asked. No task: offer `workflow interview` or a task; no Run. Otherwise: the loop.
|
|
22
22
|
|
|
23
23
|
## Loop
|
|
24
24
|
|
|
25
|
-
|
|
25
|
+
Map (not a second OS) of the 8-stage loop: Specify (stages 0-1), Make (stages 2-4), Review (stages 5-7), Deliver (stage 8). Parent = Maker + orchestrator; Devil's Advocate (DA) and Judge are fresh **siblings**, never nested or resumed. Gate failures override prose. Never edit `state.json`/`current.json`.
|
|
26
26
|
|
|
27
27
|
Tier (from `run --json`; it only rises). Do not skip Goal Contract, verification, Devil's Advocate, or Judge gates; the tier only decides who fills them.
|
|
28
28
|
- **T0** trivial (docs, typo, lint): `gate goal`, Maker, `verify`, claimed-vs-reality, `gate review`, `gate judge` (deterministic); no subagents.
|
|
29
29
|
- **T1** normal: frozen Goal, red test, `verify`, lean DA, `gate review --json`; spawn Judge only on `judgeRequired`.
|
|
30
30
|
- **T2** risky (HIGH_RISK_PATH, DOT business logic, >10 files): add `loop-risky` (blast radius, Judge on STRONG_MODEL). Unseen risk: `run --new --tier T2 --tier-reason <why>`.
|
|
31
31
|
|
|
32
|
-
0. `run --json [--mode report-only|assisted|unattended]` (
|
|
32
|
+
0. `run --json [--mode report-only|assisted|unattended]` (`RUN_DIR` = `.ai-engineering-loop/runs/<runId>/`). Do not run `sync-hosts` per run; if `hostSkills.stale`, ask the user. Read `.ai-engineering-loop/glossary.md`, plus `lessons.md`/`workflow.md` if present.
|
|
33
33
|
1. Goal Contract: `loop-goal`. 2-4. Maker: `loop-maker`, then `gate maker`. 5. Verify: `loop-verify`. 6-7. DA and Judge: `loop-review`. 8. Deliver: `loop-deliver`.
|
|
34
34
|
|
|
35
35
|
Invariants (every tier):
|
|
@@ -43,7 +43,7 @@ Invariants (every tier):
|
|
|
43
43
|
|
|
44
44
|
## Report header
|
|
45
45
|
|
|
46
|
-
When a subagent
|
|
46
|
+
When a subagent returned its own result (Grok: its own `subagent_id`):
|
|
47
47
|
|
|
48
48
|
```
|
|
49
49
|
Execution Mode: TRUE_INDEPENDENT_AGENT
|
|
@@ -7,9 +7,13 @@
|
|
|
7
7
|
|
|
8
8
|
- Never run `update apply|auto|rollback|policy` unless the user asks; the CLI applies updates itself under the user's policy (`docs runtime-safety`).
|
|
9
9
|
|
|
10
|
+
## Personal workflow
|
|
11
|
+
|
|
12
|
+
No task (`ael`, "hi", "what can you do"): offer, numbered, 1) build my workflow, 2) start a task, 3) show status. Do not create a Run or edit files. For 1: `workflow interview --json`, show the pre-filled answers with their evidence, let the user confirm or change them, write the answers to a temp file, then `workflow propose --answers <file> --json` and show the diff. Install only after the user says yes: `workflow accept --proposal <id> --by <name>`. At a no-task start or at Deliver, show at most one pending `workflow suggest` proposal. Never accept for the user. The safety backbone cannot be removed (`docs personal-workflow`).
|
|
13
|
+
|
|
10
14
|
## Studio
|
|
11
15
|
|
|
12
|
-
Stage commands (`run`, `goal`, `gate`, `verify`, `review-pack`, `deliver`, `node start|activity|complete`)
|
|
16
|
+
Stage commands (`run`, `goal`, `gate`, `verify`, `review-pack`, `deliver`, `node start|activity|complete`) stream into an open Studio; `status --json` reports `studioUrl`. Emit `npx ai-engineering-loop activity heartbeat --run <runId> --phase <phase> --message <short factual summary>` only during long manual work, never before every stage. Before asking the human, persist the question with `npx ai-engineering-loop interaction question --run <runId> --id <id> --message <text>` so it appears once in Studio.
|
|
13
17
|
|
|
14
18
|
## Budget
|
|
15
19
|
|
|
@@ -6,4 +6,5 @@
|
|
|
6
6
|
- Preview remote payloads; never include context packs, usage ledgers, raw logs, or credentials.
|
|
7
7
|
- ASSISTED: wait for explicit human approval and never infer it. Then one call: `npx ai-engineering-loop deliver --run <runId> --destination <where> --summary <text> --approve --by <name>` (approves READY approval nodes, writes `RUN_DIR/delivery.json` with `humanApproved: true`, runs `gate delivery`). Without `--approve` it refuses in ASSISTED.
|
|
8
8
|
- Long form: `node status --run <runId>`, `node approve delivery-approval --yes --by <name> --run <runId>` for each READY approval node, write `RUN_DIR/delivery.json`, then `gate delivery`. If the gate says a node is not READY, follow its prerequisite hint; never bypass approval.
|
|
9
|
-
-
|
|
9
|
+
- Commit before `deliver` (its receipt binds `HEAD`; screenshots: `--evidence AC-N=<https-url>`). Paste `RUN_DIR/receipt.md` into the PR/MR Receipt section unchanged; ❌ stays ❌ (`docs pr-receipt`).
|
|
10
|
+
- Only the delivery gate completes the run. Then run `learn` so the next Run's packs carry its lessons.
|
|
@@ -6,5 +6,6 @@
|
|
|
6
6
|
- AC is a failure table: each row has a statement, required evidence, and failure cases. Do not freeze sunny-path-only. T0 may use 1-3 rows that still list failure cases.
|
|
7
7
|
- Emit `npx ai-engineering-loop activity goal-drafting --run <runId> --message <summary>`, write the contract JSON to a repository-local temporary file, then `npx ai-engineering-loop goal draft --run <runId> --file <file> --revision <currentRevision>` so Studio shows the exact draft.
|
|
8
8
|
- Wait for the explicit human freeze, then `npx ai-engineering-loop goal freeze --latest --run <runId> --by human` (or `--revision <r> --hash <h>`). T0 may write `RUN_DIR/goal-contract.json` and run `gate goal` directly.
|
|
9
|
+
- The freeze writes a DRAFT blast-radius map: `impact show --run <runId>`. ASSISTED: the human confirms or edits it (`impact confirm --by <name> [--allow <glob>] [--remove <path>]`); copy/CSS/comment-only goals: `impact skip --by <name> --reason <text>` (`docs blast-radius-map`).
|
|
9
10
|
- Do not edit production code until the Goal gate passes.
|
|
10
11
|
- After the freeze, work from the frozen contract (`goal show --run <runId>`), not the drafting chat; compact the conversation there if the host supports it.
|
|
@@ -3,6 +3,7 @@
|
|
|
3
3
|
- T1/T2: read `docs tdd-policy` before Maker. Use `npx ai-engineering-loop context query <term>` and `context related <file>` before opening broad source files.
|
|
4
4
|
- Branch on mode. REPORT_ONLY: never edit repository files; write `RUN_DIR/report.json` with summary and recommendations, then the report review below, run `gate report`, and stop. ASSISTED: the parent (or intern) is Maker; delivery later needs human approval. UNATTENDED: `sandbox create --run <runId>`, Maker works only in the returned worktreePath, never the parent checkout, then `sandbox capture --run <runId>`.
|
|
5
5
|
- Maker intern: read `maker_intern` from workflow.md (missing or `none` = parent is Maker). Confirm it through the host model catalog; otherwise the parent is Maker and you report INVOCATION_UNAVAILABLE. The intern cannot skip later gates or spawn DA/Judge.
|
|
6
|
+
- Guardrails: Claude Code enforces them after `hooks install`. Elsewhere, before an out-of-scope edit or a destructive command (`rm -rf`, `DROP TABLE`, `git push --force`, `git reset --hard`, `git clean -fd`) run `guard check --tool edit --path <file>` or `--tool bash --command "<cmd>"`: `warn` = ask the human, `block` = do not run it.
|
|
6
7
|
- Bugs: red repro first. Features: one red test per AC row, then green.
|
|
7
8
|
- After package and manifest edits are stable, run `npx ai-engineering-loop refresh` once (drift during red-green is expected; do not refresh per test cycle).
|
|
8
9
|
- ASSISTED: `npx ai-engineering-loop gate maker --from-git` writes the exact working-tree diff (tracked and new files, excluding `.ai-engineering-loop/`) to `RUN_DIR/diff.patch` and gates it; or write that file yourself and run `gate maker`. UNATTENDED: `sandbox capture` writes it, then `gate maker`. The gate can raise the tier from the diff.
|
|
@@ -18,18 +18,18 @@ Same skill on every host. Before the first DA or Judge spawn, read your host's s
|
|
|
18
18
|
|
|
19
19
|
## Commands
|
|
20
20
|
|
|
21
|
-
`init|status|refresh|sync-hosts|generate-adapter|generate-workflow`: run `npx ai-engineering-loop <command>` (`loop-commands`); do not commit unless asked. Otherwise: the loop.
|
|
21
|
+
`init|status|refresh|sync-hosts|generate-adapter|generate-workflow`: run `npx ai-engineering-loop <command>` (`loop-commands`); do not commit unless asked. No task: offer `workflow interview` or a task; no Run. Otherwise: the loop.
|
|
22
22
|
|
|
23
23
|
## Loop
|
|
24
24
|
|
|
25
|
-
|
|
25
|
+
Map (not a second OS) of the 8-stage loop: Specify (stages 0-1), Make (stages 2-4), Review (stages 5-7), Deliver (stage 8). Parent = Maker + orchestrator; Devil's Advocate (DA) and Judge are fresh **siblings**, never nested or resumed. Gate failures override prose. Never edit `state.json`/`current.json`.
|
|
26
26
|
|
|
27
27
|
Tier (from `run --json`; it only rises). Do not skip Goal Contract, verification, Devil's Advocate, or Judge gates; the tier only decides who fills them.
|
|
28
28
|
- **T0** trivial (docs, typo, lint): `gate goal`, Maker, `verify`, claimed-vs-reality, `gate review`, `gate judge` (deterministic); no subagents.
|
|
29
29
|
- **T1** normal: frozen Goal, red test, `verify`, lean DA, `gate review --json`; spawn Judge only on `judgeRequired`.
|
|
30
30
|
- **T2** risky (HIGH_RISK_PATH, DOT business logic, >10 files): add `loop-risky` (blast radius, Judge on STRONG_MODEL). Unseen risk: `run --new --tier T2 --tier-reason <why>`.
|
|
31
31
|
|
|
32
|
-
0. `run --json [--mode report-only|assisted|unattended]` (
|
|
32
|
+
0. `run --json [--mode report-only|assisted|unattended]` (`RUN_DIR` = `.ai-engineering-loop/runs/<runId>/`). Do not run `sync-hosts` per run; if `hostSkills.stale`, ask the user. Read `.ai-engineering-loop/glossary.md`, plus `lessons.md`/`workflow.md` if present.
|
|
33
33
|
1. Goal Contract: `loop-goal`. 2-4. Maker: `loop-maker`, then `gate maker`. 5. Verify: `loop-verify`. 6-7. DA and Judge: `loop-review`. 8. Deliver: `loop-deliver`.
|
|
34
34
|
|
|
35
35
|
Invariants (every tier):
|
|
@@ -43,7 +43,7 @@ Invariants (every tier):
|
|
|
43
43
|
|
|
44
44
|
## Report header
|
|
45
45
|
|
|
46
|
-
When a subagent
|
|
46
|
+
When a subagent returned its own result (Grok: its own `subagent_id`):
|
|
47
47
|
|
|
48
48
|
```
|
|
49
49
|
Execution Mode: TRUE_INDEPENDENT_AGENT
|
|
@@ -7,9 +7,13 @@
|
|
|
7
7
|
|
|
8
8
|
- Never run `update apply|auto|rollback|policy` unless the user asks; the CLI applies updates itself under the user's policy (`docs runtime-safety`).
|
|
9
9
|
|
|
10
|
+
## Personal workflow
|
|
11
|
+
|
|
12
|
+
No task (`ael`, "hi", "what can you do"): offer, numbered, 1) build my workflow, 2) start a task, 3) show status. Do not create a Run or edit files. For 1: `workflow interview --json`, show the pre-filled answers with their evidence, let the user confirm or change them, write the answers to a temp file, then `workflow propose --answers <file> --json` and show the diff. Install only after the user says yes: `workflow accept --proposal <id> --by <name>`. At a no-task start or at Deliver, show at most one pending `workflow suggest` proposal. Never accept for the user. The safety backbone cannot be removed (`docs personal-workflow`).
|
|
13
|
+
|
|
10
14
|
## Studio
|
|
11
15
|
|
|
12
|
-
Stage commands (`run`, `goal`, `gate`, `verify`, `review-pack`, `deliver`, `node start|activity|complete`)
|
|
16
|
+
Stage commands (`run`, `goal`, `gate`, `verify`, `review-pack`, `deliver`, `node start|activity|complete`) stream into an open Studio; `status --json` reports `studioUrl`. Emit `npx ai-engineering-loop activity heartbeat --run <runId> --phase <phase> --message <short factual summary>` only during long manual work, never before every stage. Before asking the human, persist the question with `npx ai-engineering-loop interaction question --run <runId> --id <id> --message <text>` so it appears once in Studio.
|
|
13
17
|
|
|
14
18
|
## Budget
|
|
15
19
|
|
|
@@ -6,4 +6,5 @@
|
|
|
6
6
|
- Preview remote payloads; never include context packs, usage ledgers, raw logs, or credentials.
|
|
7
7
|
- ASSISTED: wait for explicit human approval and never infer it. Then one call: `npx ai-engineering-loop deliver --run <runId> --destination <where> --summary <text> --approve --by <name>` (approves READY approval nodes, writes `RUN_DIR/delivery.json` with `humanApproved: true`, runs `gate delivery`). Without `--approve` it refuses in ASSISTED.
|
|
8
8
|
- Long form: `node status --run <runId>`, `node approve delivery-approval --yes --by <name> --run <runId>` for each READY approval node, write `RUN_DIR/delivery.json`, then `gate delivery`. If the gate says a node is not READY, follow its prerequisite hint; never bypass approval.
|
|
9
|
-
-
|
|
9
|
+
- Commit before `deliver` (its receipt binds `HEAD`; screenshots: `--evidence AC-N=<https-url>`). Paste `RUN_DIR/receipt.md` into the PR/MR Receipt section unchanged; ❌ stays ❌ (`docs pr-receipt`).
|
|
10
|
+
- Only the delivery gate completes the run. Then run `learn` so the next Run's packs carry its lessons.
|
|
@@ -6,5 +6,6 @@
|
|
|
6
6
|
- AC is a failure table: each row has a statement, required evidence, and failure cases. Do not freeze sunny-path-only. T0 may use 1-3 rows that still list failure cases.
|
|
7
7
|
- Emit `npx ai-engineering-loop activity goal-drafting --run <runId> --message <summary>`, write the contract JSON to a repository-local temporary file, then `npx ai-engineering-loop goal draft --run <runId> --file <file> --revision <currentRevision>` so Studio shows the exact draft.
|
|
8
8
|
- Wait for the explicit human freeze, then `npx ai-engineering-loop goal freeze --latest --run <runId> --by human` (or `--revision <r> --hash <h>`). T0 may write `RUN_DIR/goal-contract.json` and run `gate goal` directly.
|
|
9
|
+
- The freeze writes a DRAFT blast-radius map: `impact show --run <runId>`. ASSISTED: the human confirms or edits it (`impact confirm --by <name> [--allow <glob>] [--remove <path>]`); copy/CSS/comment-only goals: `impact skip --by <name> --reason <text>` (`docs blast-radius-map`).
|
|
9
10
|
- Do not edit production code until the Goal gate passes.
|
|
10
11
|
- After the freeze, work from the frozen contract (`goal show --run <runId>`), not the drafting chat; compact the conversation there if the host supports it.
|
|
@@ -3,6 +3,7 @@
|
|
|
3
3
|
- T1/T2: read `docs tdd-policy` before Maker. Use `npx ai-engineering-loop context query <term>` and `context related <file>` before opening broad source files.
|
|
4
4
|
- Branch on mode. REPORT_ONLY: never edit repository files; write `RUN_DIR/report.json` with summary and recommendations, then the report review below, run `gate report`, and stop. ASSISTED: the parent (or intern) is Maker; delivery later needs human approval. UNATTENDED: `sandbox create --run <runId>`, Maker works only in the returned worktreePath, never the parent checkout, then `sandbox capture --run <runId>`.
|
|
5
5
|
- Maker intern: read `maker_intern` from workflow.md (missing or `none` = parent is Maker). Confirm it through the host model catalog; otherwise the parent is Maker and you report INVOCATION_UNAVAILABLE. The intern cannot skip later gates or spawn DA/Judge.
|
|
6
|
+
- Guardrails: Claude Code enforces them after `hooks install`. Elsewhere, before an out-of-scope edit or a destructive command (`rm -rf`, `DROP TABLE`, `git push --force`, `git reset --hard`, `git clean -fd`) run `guard check --tool edit --path <file>` or `--tool bash --command "<cmd>"`: `warn` = ask the human, `block` = do not run it.
|
|
6
7
|
- Bugs: red repro first. Features: one red test per AC row, then green.
|
|
7
8
|
- After package and manifest edits are stable, run `npx ai-engineering-loop refresh` once (drift during red-green is expected; do not refresh per test cycle).
|
|
8
9
|
- ASSISTED: `npx ai-engineering-loop gate maker --from-git` writes the exact working-tree diff (tracked and new files, excluding `.ai-engineering-loop/`) to `RUN_DIR/diff.patch` and gates it; or write that file yourself and run `gate maker`. UNATTENDED: `sandbox capture` writes it, then `gate maker`. The gate can raise the tier from the diff.
|
package/CHANGELOG.md
CHANGED
|
@@ -6,6 +6,75 @@ All notable changes to this project are documented here. Versions follow [Semant
|
|
|
6
6
|
|
|
7
7
|
_No changes yet._
|
|
8
8
|
|
|
9
|
+
## 2.0.0 - 2026-10-17
|
|
10
|
+
|
|
11
|
+
### Breaking changes
|
|
12
|
+
|
|
13
|
+
- Studio no longer creates or edits nodes (#80). Wiring mode only edits edges between existing nodes. `POST /api/import/preview` and `POST /api/import/apply` return 404. Author recipes as files plus `recipe create|clone|install`.
|
|
14
|
+
- `GET /api/runs/:id/lifecycle/stream` returns 404; subscribe to `GET /api/events` instead (#80).
|
|
15
|
+
- **Upgrade:** run `ai-engineering-loop sync-hosts` once after upgrading so installed host skills match 2.0.0.
|
|
16
|
+
|
|
17
|
+
### Personal workflow onboarding (#81)
|
|
18
|
+
|
|
19
|
+
- **Added:** `workflow interview|propose|accept|reject|suggest`. With no concrete task the skill offers to build a personal recipe, start a task, or show status, and does not create a Run. Interview questions come pre-filled from the last 20 run summaries (80% majority). `propose` compiles a candidate through the recipe safety backbone. `accept --by` is the only way to install it.
|
|
20
|
+
- **Added:** the personal recipe, its version history, proposals, and `workflow-history.jsonl` live under the AEL home (`AEL_HOME` or `~/.ai-engineering-loop`), never in a repository. Each terminal Run appends a summary of counts and a repo hash. Suggestions are proposals only; a rejected proposal stays quiet until new evidence outgrows the old one.
|
|
21
|
+
- **Changed:** `run` resolves project-required (`workflow.requiredRecipe`, `workflow.allowPersonal: false`, required project nodes, `adapter.md`) then the personal recipe, then the built-in default (`audit` for REPORT_ONLY). `run --json` reports `workflowSource`. `run --recipe` cannot bypass the project-required layer. UNATTENDED cannot accept (`HUMAN_REQUIRED`).
|
|
22
|
+
- **Added:** eval case `25-no-task-onboarding`; the always-loaded skill gains a no-task line (offer `workflow interview` or a task; no Run), with details in `loop-commands`.
|
|
23
|
+
- **Docs:** `core/personal-workflow.md` (`docs personal-workflow`).
|
|
24
|
+
|
|
25
|
+
### `ael bench`: compliance self-check for one host and model (#79)
|
|
26
|
+
|
|
27
|
+
- **Added:** `bench [--host <id> --model <id>] [--scenario <id>]... [--repeat <n>] [--bin <path>] [--host-arg <arg>]... [--sandbox-home] [--mode assisted|unattended|report-only] [--timeout <seconds>] [--json]`. One target per invocation. Each scenario runs in a fresh temp repo. `AEL_HOME` is a temp directory. The host process keeps the real `HOME` so it can authenticate; `--bin` and `--sandbox-home` use a temp `HOME` instead. Skills are installed in the temp repo (the directories that host reads), and `node_modules/.bin/ai-engineering-loop` shims this checkout so `npx` does not download the published package. Default timeout is 20 minutes. A missing binary is `UNAVAILABLE`, not a model failure. Default mode is ASSISTED; the scenario's scripted approver answers Goal freeze, impact confirm, and `deliver --approve`. `--mode unattended` is opt-in. Progress lines go to stderr.
|
|
28
|
+
- **Added:** `evals/scenarios/` (T0 docs, T1 features, T2 auth/payment, and traps) and `schemas/bench-scenario.schema.json`. The `evals/cases` judgment suite is unchanged by the bench.
|
|
29
|
+
- **Added:** scoring from Run artifacts only (`state.json`, `guard-events.json`, `impact-drift.json`, `verifyReceipt()`, the usage ledger, the harness clock). Tokens are parsed from host JSON when the host provides it (Claude Code and Codex). Grok CLI `-p` is plain text, so its tokens stay unknown, never `0`. `budget record --unavailable` is the same unknown. Bench-only host templates: Grok `--no-auto-update --always-approve`, Claude `--dangerously-skip-permissions`, Codex `--full-auto`. `--host-arg` appends extra flags. `--bin` overrides the binary.
|
|
30
|
+
- **Added:** `.ai-engineering-loop/bench/<benchId>/{results.json,report.md,hardening.md}` (`schemas/bench-result.schema.json`). The report is per-gate compliance, guard violations, scope drift, receipt validity, and p50/p90 time and tokens for that single target. `hardening.md` maps each violation to a `skill-src/` instruction, ranked by frequency, with a text-tightening or hard-gate suggestion. Suggestions are not applied. There is no routing or model recommendation.
|
|
31
|
+
- **Added:** `bench compare <benchIdA> <benchIdB>` prints a read-only side-by-side table.
|
|
32
|
+
- **Docs:** `core/bench.md` (`docs bench`).
|
|
33
|
+
|
|
34
|
+
### Studio: real-time sync and compact live-run UI (#80)
|
|
35
|
+
|
|
36
|
+
- **Added:** `lib/run-watcher.js` recursively watches `.ai-engineering-loop/runs/` and `current.json` (debounce ~100 ms) and pushes typed events on `GET /api/events` with monotonic `seq`. Event types: `run.created`, `run.state`, `lifecycle`, `artifact.impact|drift|guard|verification|findings|verdict|receipt`, `current.changed`, `presence`. `hello` reports `watch` or `poll` plus the watched root. `Last-Event-ID` / `?after=<seq>` replay from an in-memory ring; a gap sends `resync` for a full `/api/bootstrap`. `AEL_STUDIO_FS_WATCH=0` (and a failed `fs.watch`) uses a 1 s stat-poll.
|
|
37
|
+
- **Added:** compact live-run view as the default route (stage rail, heartbeat age, tier/mode/iteration/budget, guard/drift/receipt chips, pending-question banner, merged timeline). The Goal sheet and run-history list remain. Wiring mode edits edges between existing nodes only; `/api/validate` is still `lib/recipe.js`, and install still requires the confirm dialog.
|
|
38
|
+
- **Added:** watched-root display. `status --json` and `run --json` include `studioUrl` (tokenless Studio identity; the browser session comes from the URL printed by `ael studio`) when a Studio process for that root is alive (`.ai-engineering-loop/studio.json`).
|
|
39
|
+
- **Added:** `/api/install` requires `baseRecipeId` and rejects node-set changes (`STUDIO_STRUCTURE_READ_ONLY`); `/api/validate` reports the same code when `baseRecipeId` is sent. Only edges, `dependsOn`, and layout may differ from the loaded base recipe, including a builtin installed as `<id>-local`.
|
|
40
|
+
- **Added:** the Studio client reconnects to `/api/events` with `setTimeout` backoff and `Last-Event-ID` / `?after=`; a `resync` event or failed resume falls back to `/api/bootstrap`. The Retry button remains.
|
|
41
|
+
- **Breaking:** removed the 1.8 s `refreshLive` client poll and the 750 ms per-client lifecycle interval. `GET /api/runs/:id/lifecycle/stream` returns 404; use `GET /api/events`.
|
|
42
|
+
- **Breaking:** removed node creation, duplication, deletion, inspector field edits, loop-group creation, Custom Agent dialog, Import workflow JSON (preview/merge/replace), and the export/copy-brief toolbar. `POST /api/import/preview` and `POST /api/import/apply` return 404. Author recipes as files + `recipe create|clone|install`.
|
|
43
|
+
- **Changed:** artifact SSE summaries match real writers: `verification.json` derives `passed`/`failed` from the command bundle (no `status`); findings `blocking` uses `isBlockingFinding`; `receipt.json` verification is `verifyReceipt` (`present`, `digest`, `verified`, `problems`).
|
|
44
|
+
- **Changed:** RunWatcher processes only `current.json` and paths under `runs/`; `usage/`, `learnings.json`, drafts, and other context files emit no events.
|
|
45
|
+
|
|
46
|
+
## 1.17.0 - 2026-10-17
|
|
47
|
+
|
|
48
|
+
### `ael learn`: past Runs become project learnings for context packs (#61)
|
|
49
|
+
|
|
50
|
+
- **Added:** `learn [--dry-run] [--prune] [--max-age-runs <n>]` (`lib/learn.js`, `schemas/learnings.schema.json`). It reads every Run's `findings.json`, keeps only VALID findings that were not rejected (drift findings from #71 are skipped), clusters them per axis by word overlap, and writes deterministic, deduped `.ai-engineering-loop/learnings.json` and `learnings.md` with occurrence counts and source `runIds`. Text is redacted with the `safe-context` secret and home-path filters.
|
|
51
|
+
- **Added:** `--prune` suppresses lessons whose files are gone or that have not recurred in the newest N Runs (default 20); an entry with `supersededBy` is suppressed on the next run. A suppressed lesson returns only when it recurs in a newer Run.
|
|
52
|
+
- **Added:** Maker and Devil's Advocate context packs (`context maker|devil-advocate`, `review-pack`) include up to 5 relevant `learnings` within 10% of the stage token budget, counted in `estimatedTokens`; Judge packs get none.
|
|
53
|
+
- **Docs:** `core/learnings.md` (`docs learnings`); the Stage 8 reference runs `learn` after delivery.
|
|
54
|
+
|
|
55
|
+
### Proof-carrying PR receipt: `receipt.json` and `verify-receipt` (#70)
|
|
56
|
+
|
|
57
|
+
- **Added:** `schemas/receipt.schema.json` and `lib/receipt.js`. `deliver` (incl. `--approve`) writes `RUN_DIR/receipt.json` and `receipt.md` as its last step: the frozen Goal with its hashes, every AC with its evidence (claimed-vs-reality row, verification commands with output sha256, hosted links), the finding-ledger summary and Judge verdict, impact drift and guard counts, base/head SHA, the sha256 of every referenced artifact, and an overall digest. An AC without evidence is an explicit ❌ (`MISSING`), never dropped.
|
|
58
|
+
- **Added:** `deliver --host <id> --model <id> --head <sha> --evidence AC-N=<https-url>`. Evidence links must be `http(s)` without credentials; the receipt carries no secrets or local absolute paths.
|
|
59
|
+
- **Changed:** `delivery.json` stores `receipt: { path, digest }` (`schemas/delivery.schema.json`).
|
|
60
|
+
- **Added:** `verify-receipt [--run <id> | <receipt.json>] [--head <sha>]` re-hashes the artifacts and Goal, checks every AC has evidence, the verdict is PASS with no blocking finding, and the head SHA matches; exit 1 on any mismatch.
|
|
61
|
+
- **Changed:** the GitHub, GitLab and DOT MR templates have a Receipt section; the Stage 8 reference says to paste `receipt.md` unchanged. Spec: `core/pr-receipt.md` (`docs pr-receipt`).
|
|
62
|
+
|
|
63
|
+
### Hook-enforced runtime guardrails: scope freeze and destructive commands (#59)
|
|
64
|
+
|
|
65
|
+
- **Added:** optional `scope: { allowedPaths, deniedPaths }` (repository-relative globs) in the Goal contract, frozen with the Goal and validated by the Goal gate. Absent = no path enforcement.
|
|
66
|
+
- **Added:** `guard check --tool edit|bash` returns `allow | warn | block` with rule ids and reasons. Edits are checked against the Goal scope plus the confirmed impact map (#71) and `deniedPaths`; Bash is checked for `rm -rf`, `DROP TABLE`/`TRUNCATE`, `git push --force`, `git reset --hard` and `git clean -fd`. ASSISTED warns, UNATTENDED and REPORT_ONLY block (exit 2). Every warn/block is written to `RUN_DIR/guard-events.json`, and `gate` prints the counts (`guard` in `--json`).
|
|
67
|
+
- **Added:** `hooks install | uninstall | status` writes an idempotent Claude Code `PreToolUse` hook to the project's `.claude/settings.json` (`guard check --hook claude` answers `deny` / `ask`). Other hooks in the file are kept.
|
|
68
|
+
- **Added:** the host registry marks guardrails as **enforced** (native hooks: Claude Code) or **advisory**; shown in `hosts list`, `hosts doctor` and the README host table.
|
|
69
|
+
- **Docs:** `core/runtime-safety.md` "Runtime guardrails"; the Maker reference tells advisory hosts to call `guard check` themselves.
|
|
70
|
+
|
|
71
|
+
### Blast-radius impact map before Stage 3 (#71)
|
|
72
|
+
|
|
73
|
+
- **Added:** `schemas/impact-map.schema.json` and `lib/impact-map.js`. Freezing the Goal (`goal freeze` or `gate goal`) writes a DRAFT `RUN_DIR/impact-map.json` listing files, modules, endpoints, DB tables/migrations and tests, each with a confidence and a reason. It is a deterministic scan of the repository index against keywords from the frozen Goal (English and Indonesian filler words dropped), so it works on every adapter. `gate maker` writes it too if it is still missing.
|
|
74
|
+
- **Added:** `impact map | show | confirm | skip | drift`. `confirm --by <name>` (with optional `--allow <glob>` / `--remove <path>`) locks the map and writes `scope.allowedPaths`, the Run's allowed scope. `skip --reason <text>` records why a trivial goal (copy, CSS, comments) has no map.
|
|
75
|
+
- **Added:** after the verification gate the CLI writes `impact-drift.json` (changed files vs the map). `verify` and `review-pack` print out-of-map files, the diff-hunk pack carries `impactDrift`, and the review gate adds one `LOW` finding per out-of-map file to `findings.json` (`source: "impact-map"`). Drift warns; it does not block.
|
|
76
|
+
- **Docs:** `core/blast-radius-map.md` (`docs blast-radius-map`); the Stage 1 reference and `task-impact-inquiry` point at the map.
|
|
77
|
+
|
|
9
78
|
## 1.16.0 - 2026-10-16
|
|
10
79
|
|
|
11
80
|
### Host certification tiers and an instruction-only `AGENTS.md` digest (#60)
|
package/README.md
CHANGED
|
@@ -167,31 +167,33 @@ Under the hood, these moments use eight evidence-backed stages. The rigor is the
|
|
|
167
167
|
AI Engineering Loop is designed to work with the tools developers already use:
|
|
168
168
|
|
|
169
169
|
<!-- hosts-table:start (generated from lib/hosts/registry.js; run npm run build:hosts-table) -->
|
|
170
|
-
| Host | id | Certification | Skill folder | Reviewer agents | Detected by |
|
|
171
|
-
|
|
172
|
-
| Claude Code | `claude` | experimental | `~/.claude/skills` | `~/.claude/agents` | `~/.claude`, `claude` |
|
|
173
|
-
| Codex CLI | `codex` | experimental | `~/.agents/skills` (shared) | `~/.codex/agents` | `~/.codex`, `codex` |
|
|
174
|
-
| Gemini CLI | `gemini` | experimental | `~/.agents/skills` (shared) | `~/.gemini/agents` | `~/.gemini`, `gemini` |
|
|
175
|
-
| Antigravity (IDE / 2.0) | `antigravity` | experimental | `~/.gemini/config/skills` | `~/.gemini/config/agents` | `~/.gemini/antigravity`, `antigravity` |
|
|
176
|
-
| Antigravity CLI (agy) | `antigravity-cli` | experimental | `~/.gemini/antigravity-cli/skills` | `~/.gemini/config/agents` | `~/.gemini/antigravity-cli`, `agy` |
|
|
177
|
-
| Grok CLI | `grok` | experimental | `~/.agents/skills` (shared) | `~/.grok/agents` | `~/.grok`, `grok` |
|
|
178
|
-
| Cursor (editor + CLI) | `cursor` | experimental | `~/.agents/skills` (shared) | `~/.cursor/agents` | `~/.cursor`, `cursor-agent`, `cursor` |
|
|
179
|
-
| GitHub Copilot (CLI / VS Code) | `copilot` | experimental | `~/.agents/skills` (shared) | `~/.copilot/agents` | `~/.copilot`, `copilot` |
|
|
180
|
-
| OpenCode | `opencode` | experimental | `~/.agents/skills` (shared) | `~/.config/opencode/agents` | `~/.config/opencode`, `opencode` |
|
|
181
|
-
| Windsurf / Devin Desktop | `windsurf` | instruction-only | `~/.agents/skills` (shared) | none (in-session review) | `~/.codeium/windsurf`, `~/.config/devin` |
|
|
182
|
-
| Amp | `amp` | instruction-only | `~/.agents/skills` (shared) | none (in-session review) | `~/.config/amp`, `amp` |
|
|
183
|
-
| Factory Droid | `factory` | experimental | `~/.agents/skills` (shared) | `~/.factory/droids` | `~/.factory`, `droid` |
|
|
184
|
-
| Augment (Auggie) | `augment` | experimental | `~/.agents/skills` (shared) | `~/.augment/agents` | `~/.augment`, `auggie` |
|
|
185
|
-
| Zed | `zed` | instruction-only | `~/.agents/skills` (shared) | none (in-session review) | `~/.config/zed`, `~/Library/Application Support/Zed`, `zed` |
|
|
186
|
-
| Goose | `goose` | instruction-only | `~/.agents/skills` (shared) | none (in-session review) | `~/.config/goose`, `goose` |
|
|
187
|
-
| Cline | `cline` | instruction-only | `~/.cline/skills` | none (in-session review) | `~/.cline`, `cline` |
|
|
188
|
-
| Roo Code | `roo` | instruction-only | `~/.agents/skills` (shared) | none (in-session review) | `~/.roo` |
|
|
189
|
-
| Kiro (IDE + CLI) | `kiro` | experimental | `~/.kiro/skills` | `~/.kiro/agents` | `~/.kiro`, `kiro-cli`, `kiro` |
|
|
190
|
-
| Qwen Code | `qwen` | experimental | `~/.qwen/skills` | `~/.qwen/agents` | `~/.qwen`, `qwen` |
|
|
191
|
-
| JetBrains Junie | `junie` | experimental | `~/.agents/skills` (shared) | `~/.junie/agents` | `~/.junie`, `junie` |
|
|
170
|
+
| Host | id | Certification | Guardrails | Skill folder | Reviewer agents | Detected by |
|
|
171
|
+
|---|---|---|---|---|---|---|
|
|
172
|
+
| Claude Code | `claude` | experimental | enforced | `~/.claude/skills` | `~/.claude/agents` | `~/.claude`, `claude` |
|
|
173
|
+
| Codex CLI | `codex` | experimental | advisory | `~/.agents/skills` (shared) | `~/.codex/agents` | `~/.codex`, `codex` |
|
|
174
|
+
| Gemini CLI | `gemini` | experimental | advisory | `~/.agents/skills` (shared) | `~/.gemini/agents` | `~/.gemini`, `gemini` |
|
|
175
|
+
| Antigravity (IDE / 2.0) | `antigravity` | experimental | advisory | `~/.gemini/config/skills` | `~/.gemini/config/agents` | `~/.gemini/antigravity`, `antigravity` |
|
|
176
|
+
| Antigravity CLI (agy) | `antigravity-cli` | experimental | advisory | `~/.gemini/antigravity-cli/skills` | `~/.gemini/config/agents` | `~/.gemini/antigravity-cli`, `agy` |
|
|
177
|
+
| Grok CLI | `grok` | experimental | advisory | `~/.agents/skills` (shared) | `~/.grok/agents` | `~/.grok`, `grok` |
|
|
178
|
+
| Cursor (editor + CLI) | `cursor` | experimental | advisory | `~/.agents/skills` (shared) | `~/.cursor/agents` | `~/.cursor`, `cursor-agent`, `cursor` |
|
|
179
|
+
| GitHub Copilot (CLI / VS Code) | `copilot` | experimental | advisory | `~/.agents/skills` (shared) | `~/.copilot/agents` | `~/.copilot`, `copilot` |
|
|
180
|
+
| OpenCode | `opencode` | experimental | advisory | `~/.agents/skills` (shared) | `~/.config/opencode/agents` | `~/.config/opencode`, `opencode` |
|
|
181
|
+
| Windsurf / Devin Desktop | `windsurf` | instruction-only | advisory | `~/.agents/skills` (shared) | none (in-session review) | `~/.codeium/windsurf`, `~/.config/devin` |
|
|
182
|
+
| Amp | `amp` | instruction-only | advisory | `~/.agents/skills` (shared) | none (in-session review) | `~/.config/amp`, `amp` |
|
|
183
|
+
| Factory Droid | `factory` | experimental | advisory | `~/.agents/skills` (shared) | `~/.factory/droids` | `~/.factory`, `droid` |
|
|
184
|
+
| Augment (Auggie) | `augment` | experimental | advisory | `~/.agents/skills` (shared) | `~/.augment/agents` | `~/.augment`, `auggie` |
|
|
185
|
+
| Zed | `zed` | instruction-only | advisory | `~/.agents/skills` (shared) | none (in-session review) | `~/.config/zed`, `~/Library/Application Support/Zed`, `zed` |
|
|
186
|
+
| Goose | `goose` | instruction-only | advisory | `~/.agents/skills` (shared) | none (in-session review) | `~/.config/goose`, `goose` |
|
|
187
|
+
| Cline | `cline` | instruction-only | advisory | `~/.cline/skills` | none (in-session review) | `~/.cline`, `cline` |
|
|
188
|
+
| Roo Code | `roo` | instruction-only | advisory | `~/.agents/skills` (shared) | none (in-session review) | `~/.roo` |
|
|
189
|
+
| Kiro (IDE + CLI) | `kiro` | experimental | advisory | `~/.kiro/skills` | `~/.kiro/agents` | `~/.kiro`, `kiro-cli`, `kiro` |
|
|
190
|
+
| Qwen Code | `qwen` | experimental | advisory | `~/.qwen/skills` | `~/.qwen/agents` | `~/.qwen`, `qwen` |
|
|
191
|
+
| JetBrains Junie | `junie` | experimental | advisory | `~/.agents/skills` (shared) | `~/.junie/agents` | `~/.junie`, `junie` |
|
|
192
192
|
|
|
193
193
|
Certification: **full** = a real `default` recipe run with reviewer subagents and every gate is recorded for this host; **experimental** = reviewer subagents install, no recorded run; **instruction-only** = no subagents, review runs in-session (`CONTEXT_ISOLATION_ONLY`; compact digest: `npx ai-engineering-loop docs AGENTS`).
|
|
194
194
|
|
|
195
|
+
Guardrails: **enforced** = `npx ai-engineering-loop hooks install` adds native pre-tool hooks that check scope and destructive commands; **advisory** = the host has no hook API, so the guardrail is the loop instructions only.
|
|
196
|
+
|
|
195
197
|
Not supported (no native skills): Aider, Continue.
|
|
196
198
|
<!-- hosts-table:end -->
|
|
197
199
|
|
package/adapters/dot/gitlab.md
CHANGED
|
@@ -123,6 +123,9 @@ glab mr create \
|
|
|
123
123
|
```
|
|
124
124
|
|
|
125
125
|
#### Standardized MR Description Template:
|
|
126
|
+
|
|
127
|
+
The **Receipt** section is `RUN_DIR/receipt.md` written by `deliver` (commit first so it binds `HEAD`; screenshots as hosted links via `--evidence AC-N=<https-url>`). Paste it unchanged; a ❌ line stays ❌. Check: `npx ai-engineering-loop verify-receipt --run <runId> --head <mr-head-sha>`.
|
|
128
|
+
|
|
126
129
|
```markdown
|
|
127
130
|
## Issue
|
|
128
131
|
<GitLab Issue URL>
|
|
@@ -150,6 +153,9 @@ glab mr create \
|
|
|
150
153
|
## Screenshot / Video
|
|
151
154
|
<!-- Attach screenshots, terminal test logs, or UI captures if available -->
|
|
152
155
|
|
|
156
|
+
## Receipt
|
|
157
|
+
<paste RUN_DIR/receipt.md from `deliver` here: one ✅/❌ line per AC with its evidence>
|
|
158
|
+
|
|
153
159
|
## Note for Deployment
|
|
154
160
|
- No manual database migrations required. (or specify migration command if applicable)
|
|
155
161
|
```
|
|
@@ -29,6 +29,9 @@ gh pr create \
|
|
|
29
29
|
|
|
30
30
|
## Verification
|
|
31
31
|
Commands from `.ai-engineering-loop/verification.md` (exit 0).
|
|
32
|
+
|
|
33
|
+
## Receipt
|
|
34
|
+
<paste RUN_DIR/receipt.md from `deliver` here: one ✅/❌ line per AC with its evidence>
|
|
32
35
|
EOF
|
|
33
36
|
)"
|
|
34
37
|
```
|
|
@@ -36,3 +39,5 @@ EOF
|
|
|
36
39
|
Link an issue with `Fixes #n` in the body when an issue exists.
|
|
37
40
|
|
|
38
41
|
Do not paste raw diffs, tokens, or local machine paths.
|
|
42
|
+
|
|
43
|
+
**Receipt.** Commit the change first, then run `deliver` (it binds the receipt to `HEAD`, or pass `--head <sha>`; link hosted screenshots with `--evidence AC-N=<https-url>`). Paste `RUN_DIR/receipt.md` into the Receipt section as-is. A ❌ line is missing evidence: never delete it or flip it to ✅ by hand. Reviewers or CI check it with `npx ai-engineering-loop verify-receipt --run <runId> --head <pr-head-sha>` (exit 1 on any mismatch). Spec: `docs pr-receipt`.
|
|
@@ -31,6 +31,9 @@ glab mr create \
|
|
|
31
31
|
|
|
32
32
|
## Verification
|
|
33
33
|
Commands from `.ai-engineering-loop/verification.md` (exit 0).
|
|
34
|
+
|
|
35
|
+
## Receipt
|
|
36
|
+
<paste RUN_DIR/receipt.md from `deliver` here: one ✅/❌ line per AC with its evidence>
|
|
34
37
|
EOF
|
|
35
38
|
)" \
|
|
36
39
|
--yes
|
|
@@ -39,3 +42,5 @@ EOF
|
|
|
39
42
|
Link the issue URL in the description when an issue exists.
|
|
40
43
|
|
|
41
44
|
Do not paste raw diffs, tokens, or local machine paths.
|
|
45
|
+
|
|
46
|
+
**Receipt.** Commit the change first, then run `deliver` (it binds the receipt to `HEAD`, or pass `--head <sha>`; link hosted screenshots with `--evidence AC-N=<https-url>`). Paste `RUN_DIR/receipt.md` into the Receipt section as-is. A ❌ line is missing evidence: never delete it or flip it to ✅ by hand. Reviewers or CI check it with `npx ai-engineering-loop verify-receipt --run <runId> --head <pr-head-sha>` (exit 1 on any mismatch). Spec: `docs pr-receipt`.
|
|
@@ -33,6 +33,8 @@ Green unit tests later are not proof that siblings, approvals, or downstream job
|
|
|
33
33
|
|
|
34
34
|
Impact-only (no implementation): run this skill and stop after the report.
|
|
35
35
|
|
|
36
|
+
Inside a Run, the CLI writes a heuristic blast-radius map when the Goal is frozen (`RUN_DIR/impact-map.json`). Use it as the starting matrix, then record the outcome: `npx ai-engineering-loop impact confirm --run <runId> --by <name> [--allow <glob>]... [--remove <path>]...`. The confirmed paths become the Run's allowed scope, and verify reports changes outside it.
|
|
37
|
+
|
|
36
38
|
Commit-bound work: run this inside AEL Stage 1, then freeze. Do not run it a second time after grill.
|
|
37
39
|
|
|
38
40
|
## Host notes
|