Release → HEAD self-E2E
Internal LibrAgent workflow to turn last release..HEAD into a deduped E2E plan,
an optional GitHub tracking ticket, and delegated unattended runs with
independent verifiers.
Read references/scopes.md before acting. Default is read-only /
dry-run until the user names a scope.
Integrate — do not fork
| Need |
Use |
| Child spawn / isolation |
bundled delegate |
| Proof-before-done verifier child |
delegation-eval-loop |
| Matrix + Harbor tasks |
benchmarks/harbor/libragent-session-e2e-matrix/ |
Debug approvals / user prompts (ui__askUser, storeAs) |
LIBRAGENT_E2E_CONTROL=1 + /api/e2e-control/* (debug build only) |
| Optional external CI batch |
pnpm bench:session-e2e* (Harbor); secondary to this skill |
| Cursor HTTP harness |
.agents/skills/libr-delegate (out-of-app only) |
Hard rules
- Zero-human for every case the repo can run unattended; otherwise mark
blocked / platform-unsupported — never pass and never “needs-input”.
- No rerun-until-green; one honest attempt per case unless the user orders a rerun.
- Failures get a diagnosis, not silent retry loops.
- No fabricated tools — only names from the live tool list (
agent__*, workspace__*, etc.).
- Parent survives child stop —
agent__stopSession on children only; parent keeps orchestrating.
Outer app restart/crash tests may use a watchdog; nothing else.
- Preserve protections: release has no E2E routes; dev needs explicit opt-in;
loopback + capability header; isolated data dir — never production profile or real secrets.
- Independent verifier accepts only objective artifacts (API messages JSON, workspace files,
verifier scripts, Harbor
session_e2e_evidence.json) — not child prose alone.
- No product code fix / merge / global settings in this skill unless the user separately authorizes it.
- Creating this skill or loading it does not grant standing permission to open GitHub issues.
Known product limits (honest)
These remain outside full #2565 unattended coverage (see PR #2632):
- E2E control drives user prompts (
ui__askUser questions and storeAs /
requireUserInput secret inputs, via the shared pending-user-prompt) and
tool approvals. There is no separate interactive-shell control surface.
- Native GUI paths outside the app window are not drivable from the control plane.
- Full matrix still needs Windows or Linux desktop debug beside the app; cloud/Linux
without the app cannot claim native GUI E2E.
Mark dependent checklist rows blocked with reason; continue independent cases.
Phases
0. Confirm authorization scope
Parse the user message for one of:
| Scope |
Allowed |
read-only |
Pin SHAs, diff, read manifest, read GitHub (no writes), write local plan under .libragent/work/ |
plan |
Above + propose case list + ticket body draft (still no GitHub POST) |
github-write |
Create/update tracking issue once when user explicitly says so (e.g. “이슈 올려”, “create tracking issue”) |
execute-e2e |
Run delegated children + verifiers; requires debug app + LIBRAGENT_E2E_CONTROL=1 on isolated profile |
If ambiguous, stay in read-only and ask one question.
1. Pin baseline and HEAD
From the repo root (absolute workspacePath / workspaceOverride="current"):
python3 .agents/skills/release-head-self-e2e/scripts/pin_release_range.py \
--repo . --json > .libragent/work/release-head-self-e2e/<run-id>/pins.json
Script behavior:
- Picks latest annotated/lightweight git tag reachable as ancestor of
HEAD
(default: skips -*alpha*, -*beta*, -*rc*, -*dev* prerelease names).
- Sets
baselineSha = tag peel, headSha = HEAD, verifies ancestry; if no tag
or tag not ancestor → explicit error object (no silent fallback).
2. Map diff → E2E cases (version-aware)
- Load
benchmarks/harbor/libragent-session-e2e-matrix/coverage-manifest.json.
- Run hints (optional):
python3 .agents/skills/release-head-self-e2e/scripts/map_diff_to_cases.py \
--repo . --baseline "$(jq -r .baselineTag pins.json)" \
--manifest benchmarks/harbor/libragent-session-e2e-matrix/coverage-manifest.json \
--json > case-hints.json
- Human/agent merge: hints are suggestions only — ground each selected case in
git log / git diff paths and the issue #2565 expectation text. Write
plan.md with: case id, why in range, risk, runner (harbor-task | delegate-brief | blocked).
3. Dedupe prior tickets
- Search existing GitHub issues (title/body keywords:
release-head-self-e2e, baseline tag, #2565).
- Merge with completed evidence paths cited in comments; do not re-run
pass cases unless user requests.
- Record
dedupe.md with kept/skipped cases and links.
Requires github-write to post; otherwise output draft only.
4. Tracking ticket (optional)
Only in github-write scope. Single tracking issue per pinned baselineSha..headSha pair.
Body must include: pins, plan table, artifact dir, explicit “no secrets” policy, blocked list.
Link confirmed product failures only within user-granted write scope.
5. Execute self-E2E (execute-e2e)
Preconditions
- Debug LibrAgent (
pnpm tauri dev or debug build), disposable data dir.
LIBRAGENT_E2E_CONTROL=1. The capability token is harness-side only: the
external harness reads it from the disposable instance data dir. Agents and
child briefs never receive token material.
LIBRAGENT_E2E_SYNTHETIC_SECRET for masked-value fixtures only.
- Serialize: one case at a time on one app instance (queue file
queue.json).
- Prompt resolution scope: the only auto-resolver today is Harbor's trial
poll loop, which watches its own session. Executor children spawned via
agent__spawnSession have nothing answering their requireUserInput
prompts — route those cases (e.g. B-3 interactive echo) to the Harbor
runner, or mark them blocked with reason no in-app prompt resolver.
Parent (orchestrator)
- Load
@skill:delegate mechanics; keep workspace shared when children need repo + manifest.
- For each runnable case:
- Spawn executor child (
waitForResult=false first if parallel planning; serialize runs).
- Brief: self-contained task, exact matrix id, synthetic secret policy, deliverable = evidence paths + API excerpt pointers.
- Poll
agent__checkSession with backoff; on paused, one agent__messageToSession resume attempt.
- Spawn verifier child with no executor transcript as sole proof — pass
references/verifier-brief.md + paths to raw messages.json, workspace outputs, Harbor evidence.
- Update
progress.jsonl and per-case folder under .libragent/work/release-head-self-e2e/<run-id>/cases/<id>/.
- On unsupported driver →
blocked with reason; continue queue.
- Cleanup: executor terminate/delete after evidence harvest; verifier separate session; parent remains.
Negative controls (delegate brief or Harbor test_outputs.py patterns):
- E2E routes 404 without opt-in; wrong capability header rejected; stale execution id rejected;
synthetic secret absent from logs in verifier grep rules.
6. Closeout
summary.md: pass / fail / blocked counts, diagnosis per fail, limits section.
- If
github-write: comment on tracking issue with summary + artifact paths (no secret values).
- Do not merge or change default settings.
Artifact layout
.libragent/work/release-head-self-e2e/<run-id>/
pins.json
case-hints.json
plan.md
dedupe.md
queue.json
progress.jsonl
summary.md
cases/<case-id>/
executor-messages.json
verifier-verdict.md
evidence/...
Usage examples
Full command recipes: references/usage.md.
References
- scopes.md — authorization matrix
- usage.md — release→HEAD ticket + full self-E2E examples (bash + PowerShell)
- verifier-brief.md — independent child template
- executor-brief.md — unattended runner child template
1---2name: release-head-self-e2e3description: >4 Plan and run version-aware Session API self-E2E from the latest published release5 tag through current HEAD inside LibrAgent: pin baseline/head SHAs, map git diff to6 matrix cases, dedupe prior tracking tickets, optionally create/update a GitHub7 tracking issue when the user explicitly authorizes writes, then orchestrate parent→8 child delegates for zero-human execution and independent verification using real9 builtin tools and the debug-only E2E control plane (LIBRAGENT_E2E_CONTROL).10 Triggers: "HEAD부터 릴리스까지 E2E", "last release..HEAD E2E", "self e2e matrix",11 "#2565 자동", "릴리스 이후 회귀 E2E", "tracking ticket then self E2E".12 Not Harbor-only orchestration (Harbor is optional CI path); not Cursor libr-delegate13 unless the user runs from Cursor. Never invent tools; use agent__* and workspace__*.14metadata:15 source: https://skillmd.com/skills/fritzprix/release-head-self-e2e16---1718# Release → HEAD self-E2E1920Internal LibrAgent workflow to turn **last release..HEAD** into a **deduped E2E plan**,21an optional **GitHub tracking ticket**, and **delegated unattended runs** with22independent verifiers.2324Read [references/scopes.md](references/scopes.md) before acting. Default is **read-only /25dry-run** until the user names a scope.2627## Integrate — do not fork2829| Need | Use |30| --- | --- |31| Child spawn / isolation | bundled **`delegate`** |32| Proof-before-done verifier child | **`delegation-eval-loop`** |33| Matrix + Harbor tasks | `benchmarks/harbor/libragent-session-e2e-matrix/` |34| Debug approvals / user prompts (`ui__askUser`, `storeAs`) | `LIBRAGENT_E2E_CONTROL=1` + `/api/e2e-control/*` (debug build only) |35| Optional external CI batch | `pnpm bench:session-e2e*` (Harbor); **secondary** to this skill |36| Cursor HTTP harness | `.agents/skills/libr-delegate` (out-of-app only) |3738## Hard rules3940- **Zero-human** for every case the repo can run unattended; otherwise mark41 `blocked` / `platform-unsupported` — never `pass` and never “needs-input”.42- **No rerun-until-green**; one honest attempt per case unless the user orders a rerun.43- **Failures get a diagnosis**, not silent retry loops.44- **No fabricated tools** — only names from the live tool list (`agent__*`, `workspace__*`, etc.).45- **Parent survives child stop** — `agent__stopSession` on children only; parent keeps orchestrating.46 Outer **app restart/crash** tests may use a watchdog; nothing else.47- Preserve protections: release has **no** E2E routes; dev needs **explicit opt-in**;48 **loopback** + capability header; **isolated** data dir — never production profile or real secrets.49- **Independent verifier** accepts only objective artifacts (API messages JSON, workspace files,50 verifier scripts, Harbor `session_e2e_evidence.json`) — not child prose alone.51- **No product code fix / merge / global settings** in this skill unless the user separately authorizes it.52- Creating this skill or loading it does **not** grant standing permission to open GitHub issues.5354## Known product limits (honest)5556These remain outside full #2565 unattended coverage (see PR #2632):5758- E2E control drives **user prompts** (`ui__askUser` questions and `storeAs` /59 `requireUserInput` secret inputs, via the shared pending-user-prompt) and60 **tool approvals**. There is no separate interactive-shell control surface.61- Native GUI paths outside the app window are **not** drivable from the control plane.62- Full matrix still needs **Windows or Linux desktop debug** beside the app; cloud/Linux63 without the app cannot claim native GUI E2E.6465Mark dependent checklist rows `blocked` with reason; continue independent cases.6667## Phases6869### 0. Confirm authorization scope7071Parse the user message for one of:7273| Scope | Allowed |74| --- | --- |75| `read-only` | Pin SHAs, diff, read manifest, read GitHub (no writes), write local plan under `.libragent/work/` |76| `plan` | Above + propose case list + ticket body draft (still no GitHub POST) |77| `github-write` | Create/update tracking issue **once** when user explicitly says so (e.g. “이슈 올려”, “create tracking issue”) |78| `execute-e2e` | Run delegated children + verifiers; requires debug app + `LIBRAGENT_E2E_CONTROL=1` on isolated profile |7980If ambiguous, stay in `read-only` and ask one question.8182### 1. Pin baseline and HEAD8384From the **repo root** (absolute `workspacePath` / `workspaceOverride="current"`):8586```bash87python3 .agents/skills/release-head-self-e2e/scripts/pin_release_range.py \88 --repo . --json > .libragent/work/release-head-self-e2e/<run-id>/pins.json89```9091Script behavior:9293- Picks latest **annotated/lightweight git tag** reachable as ancestor of `HEAD`94 (default: skips `-*alpha*`, `-*beta*`, `-*rc*`, `-*dev*` prerelease names).95- Sets `baselineSha` = tag peel, `headSha` = `HEAD`, verifies ancestry; if no tag96 or tag not ancestor → explicit error object (no silent fallback).9798### 2. Map diff → E2E cases (version-aware)991001. Load `benchmarks/harbor/libragent-session-e2e-matrix/coverage-manifest.json`.1012. Run hints (optional):102103```bash104python3 .agents/skills/release-head-self-e2e/scripts/map_diff_to_cases.py \105 --repo . --baseline "$(jq -r .baselineTag pins.json)" \106 --manifest benchmarks/harbor/libragent-session-e2e-matrix/coverage-manifest.json \107 --json > case-hints.json108```1091103. **Human/agent merge**: hints are suggestions only — ground each selected case in111 `git log` / `git diff` paths and the issue #2565 expectation text. Write112 `plan.md` with: case id, why in range, risk, runner (`harbor-task` | `delegate-brief` | `blocked`).113114### 3. Dedupe prior tickets115116- Search existing GitHub issues (title/body keywords: `release-head-self-e2e`, baseline tag, `#2565`).117- Merge with completed evidence paths cited in comments; do not re-run `pass` cases unless user requests.118- Record `dedupe.md` with kept/skipped cases and links.119120**Requires `github-write`** to post; otherwise output draft only.121122### 4. Tracking ticket (optional)123124Only in **`github-write`** scope. Single tracking issue per pinned `baselineSha..headSha` pair.125Body must include: pins, plan table, artifact dir, explicit “no secrets” policy, blocked list.126Link confirmed product failures only within user-granted write scope.127128### 5. Execute self-E2E (`execute-e2e`)129130**Preconditions**131132- Debug LibrAgent (`pnpm tauri dev` or debug build), disposable data dir.133- `LIBRAGENT_E2E_CONTROL=1`. The capability token is **harness-side only**: the134 external harness reads it from the disposable instance data dir. Agents and135 child briefs never receive token material.136- `LIBRAGENT_E2E_SYNTHETIC_SECRET` for masked-value fixtures only.137- Serialize: **one case at a time** on one app instance (queue file `queue.json`).138- **Prompt resolution scope:** the only auto-resolver today is Harbor's trial139 poll loop, which watches **its own** session. Executor children spawned via140 `agent__spawnSession` have nothing answering their `requireUserInput`141 prompts — route those cases (e.g. B-3 interactive echo) to the Harbor142 runner, or mark them `blocked` with reason `no in-app prompt resolver`.143144**Parent (orchestrator)**1451461. Load `@skill:delegate` mechanics; keep workspace shared when children need repo + manifest.1472. For each runnable case:148 - Spawn **executor** child (`waitForResult=false` first if parallel planning; **serialize** runs).149 - Brief: self-contained task, exact matrix id, synthetic secret policy, deliverable = evidence paths + API excerpt pointers.150 - Poll `agent__checkSession` with backoff; on `paused`, one `agent__messageToSession` resume attempt.1513. Spawn **verifier** child with **no executor transcript** as sole proof — pass152 `references/verifier-brief.md` + paths to raw `messages.json`, workspace outputs, Harbor evidence.1534. Update `progress.jsonl` and per-case folder under `.libragent/work/release-head-self-e2e/<run-id>/cases/<id>/`.1545. On unsupported driver → `blocked` with reason; continue queue.1556. Cleanup: executor terminate/delete after evidence harvest; verifier separate session; parent remains.156157**Negative controls** (delegate brief or Harbor `test_outputs.py` patterns):158159- E2E routes 404 without opt-in; wrong capability header rejected; stale execution id rejected;160 synthetic secret absent from logs in verifier grep rules.161162### 6. Closeout163164- `summary.md`: pass / fail / blocked counts, diagnosis per fail, limits section.165- If `github-write`: comment on tracking issue with summary + artifact paths (no secret values).166- Do not merge or change default settings.167168## Artifact layout169170```171.libragent/work/release-head-self-e2e/<run-id>/172 pins.json173 case-hints.json174 plan.md175 dedupe.md176 queue.json177 progress.jsonl178 summary.md179 cases/<case-id>/180 executor-messages.json181 verifier-verdict.md182 evidence/...183```184185## Usage examples186187Full command recipes: [references/usage.md](references/usage.md).188189## References190191- [scopes.md](references/scopes.md) — authorization matrix192- [usage.md](references/usage.md) — release→HEAD ticket + full self-E2E examples (bash + PowerShell)193- [verifier-brief.md](references/verifier-brief.md) — independent child template194- [executor-brief.md](references/executor-brief.md) — unattended runner child template