@ps-neko/nekowork 0.1.0-alpha.0 → 0.1.0-alpha.2

package/CLAUDE.md

@@ -8,15 +8,16 @@
 
 ## 자동 갱신 영역
 
-<!-- HARNESS:START version=0.1.0-alpha.0 -->
+<!-- HARNESS:START version=0.1.0-alpha.2 -->
 <!-- 이 영역은 scripts/sync-claude-md.js 가 자동 갱신한다. 직접 편집 금지. -->
 
 ## 카탈로그 요약
 
 - agents: 11
-- skills: 9
+- skills: 10
 - commands: 1 (legacy compat)
 - hooks: 5 (gateguard-fact-force, config-protection, quality-gate, pre-bash-dispatcher, persistent-mode)
+- packs: core, quality, security, frontend, testing, release, enterprise
 - profiles: core, developer, security, product, quality, frontend, testing, research, full
 - harnesses: claude, codex, cursor, gemini, opencode
 
@@ -40,12 +41,14 @@
 
 ```bash
 harness install --plan --profile core      # 설치 dry-run
+harness install --plan --pack quality      # curated pack dry-run
 harness ask "<task>"                       # question gate, no project mutation
 harness team "<task>"                      # read-only worker handoffs
 harness work "<task>"                      # single executor implement handoff
 harness verify "<task>" --session <id>     # Codex-only verification
 harness gate status --session <id>         # inspect or resolve HUMAN_GATE state
 harness ship "<task>" --session <id>       # ship/no-ship readiness handoff
+harness report --session <id>              # readable evidence report
 harness apply --session <id>               # apply verified SHIP_READY live-work diff
 harness run "<task>" --session <id>        # work -> verify -> ship, optional --apply
 harness review "<task>" [--secure|--fast|--no-ship]  # legacy full cycle

package/README.md

@@ -18,12 +18,54 @@ NEKOWORK = Claude work -> Codex verification -> Human Gate
 
 NEKOWORK is not meant to become a large agent pack. Skills, hooks, profiles, and team modes are added only when they preserve the verification loop.
 
+NEKOWORK intentionally keeps the catalog selective. Every agent, skill, hook, profile, module, and pack must preserve the verification loop.
+
+**Public alpha evidence:** 7 packs / 9 profiles / 36 components / 5 harness targets / 6 case-study flows / 245 tests / 0 moderate+ npm audit issues / fresh `npx @alpha` smoke
+
+NEKOWORK does not automatically commit, push, publish, deploy, or apply diffs. `apply` is explicit and requires verified ship-ready evidence.
+
+**One-minute demo:** [terminal transcript](docs/DEMO.md#one-minute-terminal-transcript) / [full report example](docs/DEMO-REPORT.md) / [alpha feedback](https://github.com/Ps-Neko/NEKOWORK/issues/new?template=alpha-feedback.yml) / [roadmap](docs/ROADMAP.md)
+
+![NEKOWORK one-minute terminal demo](docs/assets/demo-terminal.svg)
+
+## Example Report
+
+`report` is the main trust surface. It turns session evidence into a readable `REPORT.md`:
+
+```text
+Verdict: approve_with_fixes
+Ship ready: false
+Human gate: required
+Applied: false
+Profile: quality
+Strict quality: enabled
+Acceptance coverage: 4/5
+Quality warnings: 2
+
+Evidence:
+- work-summary.json
+- verify-summary.json
+- ship-summary.json
+- gate-summary.json
+```
+
+See the full report contract and example artifact in [docs/DEMO-REPORT.md](docs/DEMO-REPORT.md), and the one-minute terminal transcript in [docs/DEMO.md](docs/DEMO.md).
+
+## Compared With Agent Packs
+
+| Tool pattern | Optimizes for | NEKOWORK optimizes for |
+|---|---|---|
+| Large Claude Code packs | More agents, commands, skills | Curated verification loop |
+| Team simulation | More specialist perspectives | Read-only team plus one executor |
+| Autopilot | Fast autonomous execution | Report, gate, explicit apply |
+| Discipline workflows | Better development habits | Evidence-backed ship decision |
+
 ## Three Paths
 
 Most users should start with the Beginner path. The other paths are for explicit phase control or legacy compatibility.
 
-1. Beginner: `doctor -> ask -> run -> gate`
-2. Advanced: `ask -> plan -> team -> work -> verify -> gate -> ship -> apply`
+1. Beginner: `doctor -> ask -> run -> report -> gate`
+2. Advanced: `ask -> plan -> team -> work -> verify -> gate -> ship -> report -> apply`
 3. Legacy: `review` / `review-cycle`
 
 ## Why NEKOWORK
@@ -32,19 +74,43 @@ NEKOWORK is for teams that want AI-assisted development without making the agent
 
 ## Status
 
-- Current version: `0.1.0-alpha.0` public alpha candidate
+- Current repository version: `0.1.0-alpha.2`
 - Current package name: `@ps-neko/nekowork`
-- npm publishing: prepared for `npm publish --access public --tag alpha`, but not published until npm owner auth is available
-- Supported install path today: clone, submodule, or local repository integration
-- Future npm path is prepared; final publish requires `npm whoami` to succeed
+- Current npm alpha: `@ps-neko/nekowork@0.1.0-alpha.2`
+- Supported install path today: npm alpha, clone, submodule, or local repository integration
+- Dist-tag note: use `@alpha` until a stable release; `latest` still points at the first alpha line
 - Default mode: mock providers, no API keys, no provider CLI calls
 
 Current local verification:
 
 - `npm run lint`: pass
-- `npm test`: 238 tests pass
+- `npm test`: 245 tests pass
 - `npm audit --audit-level=moderate`: 0 vulnerabilities
 - `npm pack --dry-run --json`: pass
+- `npx -y @ps-neko/nekowork@alpha doctor --quick`: pass with warnings only
+
+## Case-study Evidence
+
+| Flow | Risk type | Evidence produced |
+|---|---|---|
+| Financial UI mock | UI/product risk | report + Human Gate |
+| GitHub Actions hardening | CI/security risk | security findings + no-ship/ship evidence |
+| Quality lifecycle smoke | quality risk | strict-quality + acceptance coverage |
+| npm package boundary | package/release risk | pack/audit evidence |
+| Auth parser boundary | auth/security risk | parser boundary evidence |
+| Python protocol parser | protocol correctness risk | test-backed verification |
+
+## Official Packs
+
+| Pack | Adds | Use when |
+|---|---|---|
+| `core` | minimal verification runtime | first install or repo smoke |
+| `quality` | acceptance coverage, strict evidence prompts | feature work needs proof |
+| `security` | auth/secrets/deploy risk prompts | sensitive changes |
+| `frontend` | UI mockup, component review, accessibility checks | product-facing UI work |
+| `testing` | regression planning and coverage handoffs | test confidence is the main risk |
+| `release` | ship/no-ship evidence | pre-release checks |
+| `enterprise` | full catalog with all gates | high-control teams |
 
 ## Quick Start
 
@@ -56,6 +122,12 @@ Requirements:
 
 Fastest no-API demo:
 
+```bash
+npx -y @ps-neko/nekowork@alpha doctor --quick
+```
+
+Repository demo:
+
 ```bash
 git clone https://github.com/Ps-Neko/NEKOWORK.git harness
 cd harness
@@ -63,7 +135,7 @@ npm ci
 npm run demo:quick -- --cleanup
 ```
 
-This creates a disposable target project and runs `doctor -> run -> gate status`. It uses mock providers and does not call Claude, Codex, Gemini, or paid APIs.
+This creates a disposable target project and runs `doctor -> run -> report -> gate status`. It uses mock providers and does not call Claude, Codex, Gemini, or paid APIs.
 
 Recommended path for most users:
 
@@ -74,15 +146,16 @@ npm ci
 node scripts/cli.js doctor --quick
 node scripts/cli.js ask "clarify a risky or ambiguous request" --session first-ask
 node scripts/cli.js run "implement, verify, and prepare ship readiness" --session first-run
+node scripts/cli.js report --session first-run
 node scripts/cli.js gate status --session first-run
 ```
 
-`run` executes `work -> verify -> ship`. It does not apply by default. `apply` is always explicit and requires a verified `SHIP_READY` live-work diff.
+`run` executes `work -> verify -> ship`. `report` turns the session evidence into a readable `REPORT.md`. It does not apply by default. `apply` is always explicit and requires a verified `SHIP_READY` live-work diff.
 
 Advanced path:
 
 ```text
-ask -> plan -> team -> work -> verify -> gate -> ship -> apply
+ask -> plan -> team -> work -> verify -> gate -> ship -> report -> apply
 ```
 
 Legacy compatibility smoke:
@@ -101,13 +174,14 @@ To see the repository-based external project flow end to end:
 npm run demo:external
 ```
 
-To inspect small case-study targets, see [examples/trading-dashboard-mock](examples/trading-dashboard-mock), [examples/github-actions-hardening](examples/github-actions-hardening), and [examples/quality-lifecycle-smoke](examples/quality-lifecycle-smoke). They demonstrate financial UI, CI workflow, and quality lifecycle changes passing local checks while still preserving Codex verification, Human Gate policy, and explicit apply control.
+To inspect small case-study targets, see [examples/trading-dashboard-mock](examples/trading-dashboard-mock), [examples/github-actions-hardening](examples/github-actions-hardening), [examples/quality-lifecycle-smoke](examples/quality-lifecycle-smoke), and [docs/case-studies](docs/case-studies). They demonstrate financial UI, CI workflow, quality lifecycle, npm package, auth parser, and Python protocol library flows passing local checks while still preserving Codex verification, Human Gate policy, and explicit apply control.
 
 ## What You Get
 
 ```text
 doctor ... OK
 run workflow ... OK
+report ... OK
 gate status ... OK
 Demo completed: verdict=approve_with_fixes, ship_ready=false, applied=false
 ```
@@ -116,6 +190,7 @@ Outputs are written under:
 
 ```text
 .harness/state/sessions/<session-id>/handoffs/
+.harness/state/sessions/<session-id>/REPORT.md
 ```
 
 ## Use It In Another Project
@@ -169,6 +244,7 @@ The public alpha surface is intentionally small:
 - `ship`: produce a ship/no-ship readiness handoff after Codex verification
 - `apply`: apply a verified `SHIP_READY` live-work diff to the target project
 - `run`: execute the decomposed wrapper, `work -> verify -> ship`, with optional apply
+- `report`: summarize session evidence into `REPORT.md` without project mutation
 - `review`: run the legacy full Claude-led/Codex-reviewed workflow
 - `review-cycle`: explicit compatibility alias for the legacy full review workflow
 - `install --plan` / `install --apply`: project generated harness surfaces
@@ -179,13 +255,24 @@ Advanced features such as `team-lite`, `ralph`, `wait`, instincts, cost tracking
 
 Use `--profile quality` or `--profile security` on `work`, `verify`, and `run` when a task needs stronger evidence prompts. Add `--strict-quality` to `verify` or `run` when missing evidence or acceptance coverage should become a fix-required verdict before ship.
 
+Use official packs when choosing an install shape:
+
+```bash
+node scripts/install-plan.js --list
+node scripts/install-plan.js --pack quality
+node scripts/install-plan.js --pack security --target codex --json
+```
+
+Packs are aliases over validated profiles. They add clearer product packaging without weakening the core gates.
+
 ## Catalog
 
 - Agents: 11
-- Skills: 9
+- Skills: 10
 - Hooks: 5
 - Modules: 7
 - Profiles: `core`, `developer`, `security`, `product`, `quality`, `frontend`, `testing`, `research`, `full`
+- Official packs: `core`, `quality`, `security`, `frontend`, `testing`, `release`, `enterprise`
 - Harness targets: `claude`, `codex`, `cursor`, `gemini`, `opencode`
 
 Key skills:
@@ -193,6 +280,7 @@ Key skills:
 - `claude-led-codex-review`
 - `plan-eng-review`
 - `tdd-workflow`
+- `acceptance-coverage`
 - `review`
 - `ship`
 - `ralph`
@@ -207,6 +295,7 @@ node scripts/cli.js doctor
 node scripts/cli.js doctor --quick --gemini-smoke
 npm run demo:quick
 node scripts/install-plan.js --list
+node scripts/install-plan.js --pack quality
 node scripts/install-plan.js --profile developer
 node scripts/install-apply.js --profile developer --project-root <target>
 
@@ -218,8 +307,10 @@ node scripts/cli.js verify "verify the implemented change" --session work-smoke
 node scripts/cli.js verify "verify quality evidence" --profile quality --strict-quality --session work-smoke
 node scripts/cli.js gate status --session work-smoke
 node scripts/cli.js ship "prepare ship readiness" --require-clean-gates --session work-smoke
+node scripts/cli.js report --session work-smoke
 node scripts/cli.js apply --session work-smoke
 node scripts/cli.js run "implement, verify, and prepare ship readiness" --session run-smoke
+node scripts/cli.js report --session run-smoke
 node scripts/cli.js review "implement and review this change" --no-ship
 node scripts/cli.js review-cycle "legacy full-cycle compatibility smoke" --no-ship
 node scripts/cli.js review "security-sensitive change" --secure --no-ship
@@ -247,14 +338,18 @@ npm run security:hardening
 npm pack --dry-run --json
 ```
 
-`npm pack --dry-run --json` currently produces a package named like `ps-neko-nekowork-0.1.0-alpha.0.tgz`. It does not publish.
+`npm pack --dry-run --json` currently produces a package named like `ps-neko-nekowork-0.1.0-alpha.2.tgz`. It does not publish.
 
 ## Documentation
 
 - [docs/QUICKSTART.md](docs/QUICKSTART.md) - first run and common paths
 - [docs/WHY-NEKOWORK.md](docs/WHY-NEKOWORK.md) - comparison and product positioning
+- [docs/CATALOG-PACKS.md](docs/CATALOG-PACKS.md) - curated catalog, official packs, and case-study evidence
 - [docs/PUBLISH-ALPHA.md](docs/PUBLISH-ALPHA.md) - public npm alpha release plan
+- [docs/ROADMAP.md](docs/ROADMAP.md) - small alpha roadmap and non-goals
+- [docs/INTERNAL-PROVIDER.md](docs/INTERNAL-PROVIDER.md) - private command adapter protocol
 - [docs/DEMO.md](docs/DEMO.md) - sample command output and generated files
+- [docs/DEMO-REPORT.md](docs/DEMO-REPORT.md) - readable session report UX
 - [docs/EXAMPLE-PROJECT.md](docs/EXAMPLE-PROJECT.md) - repository-based external project demo
 - [docs/case-studies](docs/case-studies) - real external project run evidence
 - [examples/trading-dashboard-mock](examples/trading-dashboard-mock) - standalone financial UI mock target and case-study evidence

package/agent.yaml

@@ -1,7 +1,7 @@
 spec_version: gitagent/0.1.0
 name: nekowork
 runtime_name: harness
-version: 0.1.0-alpha.0
+version: 0.1.0-alpha.2
 description: "NEKOWORK HARNESS - Local-first multi-AI development verification runtime"
 license: MIT
 homepage: https://github.com/Ps-Neko/NEKOWORK
@@ -26,6 +26,7 @@ skills:
   - claude-led-codex-review
   - plan-eng-review
   - tdd-workflow
+  - acceptance-coverage
   - review
   - ship
   - security-hardening
@@ -97,7 +98,7 @@ profiles:
     - full
 
 modules:
-  # 0.0.3 catalog: 7 modules. Future modules stay selective/profile-driven.
+  # Current catalog: 7 modules. Future modules stay selective/profile-driven.
   - rules-core
   - agents-core
   - hooks-runtime

package/docs/ADVANCED.md

@@ -174,7 +174,7 @@ Rules:
 Policy:
 
 - `run` is the short safe wrapper for new users.
-- `run` does not call `plan` in the `0.0.3` line.
+- `run` does not call `plan` in the current alpha line.
 - `plan` is recommended before `work` for larger changes.
 - `work` still records `acceptance-criteria.json`, so `run` preserves success criteria evidence.
 - `apply` is always explicit; use `run --apply` only after live work can produce a captured diff.
@@ -184,6 +184,28 @@ Outputs:
 - `.harness/state/sessions/<id>/run-summary.json`
 - all normal `work`, `verify`, `ship`, and optional `apply` outputs
 
+## report
+
+`report` turns existing session evidence into a readable inspect-only report:
+
+```bash
+node scripts/cli.js report --session run-smoke
+node scripts/cli.js report --session run-smoke --stdout
+node scripts/cli.js report --session run-smoke --output docs/session-report.md
+```
+
+Rules:
+
+- Reads summaries, markers, acceptance criteria, and handoffs from `.harness/state/sessions/<id>/`.
+- Writes `REPORT.md` and `report-summary.json` by default.
+- Does not call providers, run git commands, apply diffs, or mutate target project files.
+- Can run after `ask`, `work`, `verify`, `ship`, `run`, or `apply`.
+
+Outputs:
+
+- `.harness/state/sessions/<id>/REPORT.md`
+- `.harness/state/sessions/<id>/report-summary.json`
+
 ## review-cycle
 
 `review-cycle` is the explicit compatibility alias for the legacy full workflow:
@@ -194,7 +216,7 @@ node scripts/cli.js review-cycle "legacy full-cycle smoke" --no-ship
 
 Rules:
 
-- It is equivalent to `review` in the `0.0.3` line.
+- It is equivalent to `review` in the current alpha line.
 - It keeps the old `ideate -> plan -> implement -> self-review -> codex-review -> codex-challenge -> ship` behavior discoverable while new automation migrates to `run` or the decomposed commands.
 - It writes `review-summary.json` with `mode: legacy-full-review-cycle`.
 - It may use legacy live-review behavior, so new controlled project mutation should prefer `work --live -> verify -> ship -> apply`.
@@ -280,7 +302,7 @@ node scripts/cli.js instincts ready --blocked
 node scripts/cli.js instincts promote <id>
 ```
 
-Promotion requires confidence `1.0`; automatic promotion without human confirmation is outside the 0.0.3 release scope.
+Promotion requires confidence `1.0`; automatic promotion without human confirmation is outside the current alpha release scope.
 
 ## Cost Tracking
 
@@ -304,7 +326,7 @@ Verify it with:
 npm run verify:runtime
 ```
 
-The Node CLI remains the primary 0.0.3 user path.
+The Node CLI remains the primary alpha user path.
 
 ## Full Builder Surface
 

package/docs/AI-DEVELOPMENT-LIFECYCLE.md

@@ -54,10 +54,11 @@ ask
   -> verify
   -> gate
   -> ship
+  -> report
   -> apply
 ```
 
-Quality enters early through `ask` and `plan`, not only at the final review step. Team mode collects multiple perspectives, but the write phase stays single-executor. Verification is independent, gate decisions are explicit, and apply requires evidence.
+Quality enters early through `ask` and `plan`, not only at the final review step. Team mode collects multiple perspectives, but the write phase stays single-executor. Verification is independent, gate decisions are explicit, `report` makes evidence readable, and apply requires evidence.
 
 ## Quality Profile
 

package/docs/ARCHITECTURE.md

@@ -50,7 +50,7 @@ User command
         |
         |-- doctor
         |-- install plan/apply
-        |-- ask / plan / team / work / verify / gate / ship / apply / run / review / review-cycle
+        |-- ask / plan / team / work / verify / gate / ship / report / apply / run / review / review-cycle
         |-- ralph
         |-- team-lite
         |-- sessions / costs / instincts
@@ -74,6 +74,7 @@ The public alpha surface is intentionally small:
 
 ```bash
 node scripts/cli.js doctor
+node scripts/cli.js install --plan --pack quality
 node scripts/cli.js install --plan --profile developer
 node scripts/cli.js install --apply --profile developer --project-root <target>
 node scripts/cli.js ask "clarify a risky or ambiguous request" --project-root <target>
@@ -83,6 +84,7 @@ node scripts/cli.js work "single executor implementation" --session work-smoke -
 node scripts/cli.js verify "Codex verification" --session work-smoke --project-root <target>
 node scripts/cli.js gate status --session work-smoke --project-root <target>
 node scripts/cli.js ship "ship readiness" --session work-smoke --project-root <target>
+node scripts/cli.js report --session work-smoke --project-root <target>
 node scripts/cli.js apply --session work-smoke --project-root <target>
 node scripts/cli.js run "decomposed wrapper" --session run-smoke --project-root <target>
 node scripts/cli.js review "change request" --no-ship --project-root <target>
@@ -99,7 +101,7 @@ Advanced features are documented separately:
 
 ## Review Pipeline
 
-The `0.0.3` `review` command remains the Claude-led and Codex-reviewed legacy full cycle. `review-cycle` is an explicit compatibility alias for the same behavior:
+The current alpha `review` command remains the Claude-led and Codex-reviewed legacy full cycle. `review-cycle` is an explicit compatibility alias for the same behavior:
 
 ```text
 ideate
@@ -114,10 +116,10 @@ ideate
 The long-term phase model is additive and keeps `review` compatibility during migration:
 
 ```text
-ask -> plan -> team -> work -> verify -> gate -> ship -> apply
+ask -> plan -> team -> work -> verify -> gate -> ship -> report -> apply
 ```
 
-`ask` is a local question gate. `team` creates read-only handoffs from multiple worker perspectives. `work` lets one executor produce an implement handoff and, in live mode, an isolated workspace diff. `verify` runs Codex-only verification against that prior work handoff. `gate` records explicit human approve/block decisions for `HUMAN_GATE`. `ship` creates a ship/no-ship readiness handoff and refuses to bypass unresolved gates. `apply` is the only decomposed command in this chain that mutates the target project, and only by applying a verified `SHIP_READY` live-work diff. `team-lite` remains an advanced read-only staged handoff experiment. Future `review` can be retired or kept as a compatibility wrapper once callers have migrated to the decomposed commands.
+`ask` is a local question gate. `team` creates read-only handoffs from multiple worker perspectives. `work` lets one executor produce an implement handoff and, in live mode, an isolated workspace diff. `verify` runs Codex-only verification against that prior work handoff. `gate` records explicit human approve/block decisions for `HUMAN_GATE`. `ship` creates a ship/no-ship readiness handoff and refuses to bypass unresolved gates. `report` summarizes existing session evidence without mutating project files. `apply` is the only decomposed command in this chain that mutates the target project, and only by applying a verified `SHIP_READY` live-work diff. `team-lite` remains an advanced read-only staged handoff experiment. Future `review` can be retired or kept as a compatibility wrapper once callers have migrated to the decomposed commands.
 
 `work` does not run Codex review or ship. It also does not mutate the target project directly; live executor changes are captured as a session diff for later verification.
 
@@ -129,6 +131,8 @@ ask -> plan -> team -> work -> verify -> gate -> ship -> apply
 
 `ship` does not implement, verify, publish, deploy, or mutate the target project. It requires both prior `work` and Codex verification handoffs. It writes `SHIP_READY` only for fully approved verification or explicit human gate approval, writes `NO_SHIP` for fixable findings, and stops with a human gate when `HUMAN_GATE` is unresolved or explicitly blocked.
 
+`report` does not implement, verify, ship, apply, call providers, or inspect project source. It reads session summaries, markers, acceptance criteria, and handoffs, then writes `REPORT.md` and `report-summary.json` under the session directory.
+
 `apply` requires `SHIP_READY`, no newer `NO_SHIP`, no unresolved gate, and a captured diff from `work --live`. It applies that diff with `git apply --3way`, records `APPLIED_DIFF`, and leaves commit/push/release actions to the human.
 
 `run` is the compatibility-friendly wrapper around the decomposed path. It runs `work -> verify -> ship` and only runs `apply` when `--apply` is explicitly requested and `SHIP_READY` exists. New automation should prefer `run` or the explicit decomposed commands; old automation can continue to use `review` or `review-cycle`.
@@ -198,8 +202,8 @@ Builders project the catalog into tool-specific files:
 
 ## Release State
 
-The current release line is `0.1.0-alpha.0`:
+The current release line is `0.1.0-alpha.2`:
 
 - Repository and GitHub tarball release are available.
-- Public npm metadata is prepared, but publish execution is blocked until npm owner auth is available.
-- Clone, submodule, and local checkout integration remain the supported install paths until the package is published.
+- Public npm alpha is published as `@ps-neko/nekowork@alpha`.
+- Clone, submodule, and local checkout integration remain supported for repository-pinned workflows.

package/docs/AUDIT.md

@@ -1,25 +1,28 @@
 # Audit
 
-Status date: 2026-05-07
+Status date: 2026-05-08
 
-This audit summarizes the current NEKOWORK state after the `v0.0.3` repository release. It replaces the older week-by-week scratch audit, which contained stale planning notes and encoding damage.
+This audit summarizes the current NEKOWORK state after publishing the `0.1.0-alpha.2` public alpha. It replaces the older week-by-week scratch audit, which contained stale planning notes and encoding damage.
 
 ## Current Status
 
 | Area | Status | Notes |
 |---|---|---|
-| Package metadata | OK | `@ps-neko/nekowork@0.1.0-alpha.0`, `agent.yaml` uses `name: nekowork`, `runtime_name: harness` |
-| npm publish | Blocked on auth | Public alpha metadata is prepared; `npm whoami` currently returns `ENEEDAUTH` |
+| Package metadata | OK | `@ps-neko/nekowork@0.1.0-alpha.2`, `agent.yaml` uses `name: nekowork`, `runtime_name: harness` |
+| npm publish | OK | `@ps-neko/nekowork@alpha` points at `0.1.0-alpha.2` |
 | Source install | OK | Clone, local checkout, and submodule workflows are documented |
-| Public npm alpha plan | OK | `docs/PUBLISH-ALPHA.md` defines the `0.1.0-alpha.0` path; npm publish has not been executed because npm owner auth is unavailable |
+| Public npm alpha | OK | `docs/PUBLISH-ALPHA.md` records the first alpha publish and the `0.1.0-alpha.2` alpha update |
 | CLI doctor | OK | `doctor`, `doctor --quick`, and `doctor --gemini-smoke` are available |
 | Provider auth | OK | Local delegated CLI auth is the default path |
-| Catalog | OK | 11 agents, 9 skills, 5 hooks, 7 modules, 35 components, 9 profiles |
+| Internal provider adapter | OK | `HARNESS_PROVIDER_OVERRIDE=internal` can call an explicit JSON command adapter without weakening gates |
+| Catalog | OK | 7 official packs, 11 agents, 10 skills, 5 hooks, 7 modules, 36 components, 9 profiles |
 | Multi-harness output | OK | Claude, Codex, Cursor, Gemini, and OpenCode builders are present |
-| Quick demo | OK | `npm run demo:quick` verifies the shortest no-API `doctor -> run -> gate status` path |
+| Quick demo | OK | `npm run demo:quick` verifies the shortest no-API `doctor -> run -> report -> gate status` path |
+| Fresh npm alpha smoke | OK | CI runs `npx -y @ps-neko/nekowork@alpha doctor --quick --json` from a disposable directory |
+| Report UX | OK | `report` writes inspect-only `REPORT.md` and `report-summary.json` from session evidence |
 | External demo | OK | `npm run demo:external` verifies a disposable target project flow |
-| Third-party case study | OK | `docs/case-studies/SINDRESORHUS-IS-PLAIN-OBJ.md` records a real public repository run |
-| Decomposed workflow | OK | `ask`, `team`, `work`, `verify`, `gate`, `ship`, `apply`, and `run` are available |
+| Third-party case studies | OK | `docs/case-studies/` records real public repository runs for npm package, auth boundary, and Python protocol targets |
+| Decomposed workflow | OK | `ask`, `team`, `work`, `verify`, `gate`, `ship`, `report`, `apply`, and `run` are available |
 | Risk policy | OK | Shared classifier drives ask, routing traces, verify challenge/gates, and ship gate rechecks |
 | Acceptance criteria | OK | `work` ensures every session has `acceptance-criteria.json` |
 | Profile safety | OK | Manifest/catalog validators reject profiles that weaken core gates |
@@ -27,7 +30,7 @@ This audit summarizes the current NEKOWORK state after the `v0.0.3` repository r
 | Persistent wakeup | OK | `wait` resumes supported active sessions and blocks on `HUMAN_GATE` |
 | Generated docs | OK | CODEMAP output is stable ASCII and reproducible |
 | Tests | OK | Unit, integration, and e2e suites pass locally and in CI |
-| Release | OK | `v0.0.3` prerelease exists with tarball asset |
+| Release | OK | `v0.1.0-alpha.2` is tagged and published as a GitHub prerelease |
 
 ## Verification Gates
 
@@ -52,7 +55,7 @@ Current local result for this working tree:
 - `npm run test:unit`: covered by full `npm test`
 - `npm run validate:all`: pass
 - `npm run lint`: pass
-- `npm test`: 238 tests pass
+- `npm test`: 245 tests pass
 - quick run demo: pass through `npm run demo:quick -- --cleanup`
 - external project e2e smoke: pass through `npm test`
 - `node scripts/sync-claude-md.js --check`: pass
@@ -60,11 +63,15 @@ Current local result for this working tree:
 - `npm audit --audit-level=moderate`: 0 vulnerabilities
 - `npm pack --dry-run --json`: pass
 - `npm publish --dry-run --access public --tag alpha`: pass
-- `npm publish --access public --tag alpha`: blocked by `ENEEDAUTH`
+- `npm publish --access public --tag alpha`: `0.1.0-alpha.2` published
+- `npm view @ps-neko/nekowork dist-tags version versions --json`: `alpha` points at `0.1.0-alpha.2`; `latest` remains `0.1.0-alpha.0`
+- `npx -y @ps-neko/nekowork@alpha doctor --quick`: passed for `0.1.0-alpha.2` with WARN summary from non-git project root and Gemini auth not checked
 
 ## Completed Work
 
 - Local-first provider auth policy implemented and documented.
+- Internal provider command adapter implemented and documented without bypassing verification, Human Gate, or apply controls.
+- `acceptance-coverage` skill added as a focused quality evidence helper.
 - API-key override warnings and guards are in place.
 - Provider CLI path trust checks are in place.
 - `--project-root` separates NEKOWORK tool root from target project root.
@@ -82,22 +89,23 @@ Current local result for this working tree:
 - Release docs, setup docs, runbook, quickstart, porting guide, and CODEMAP docs are readable for external users.
 - The disposable external project demo proves the repository-based target-project flow end to end.
 - The quick run demo proves the one-command no-API first experience.
+- `report` gives public alpha users a readable inspect-only session artifact without applying or mutating project files.
+- Official packs expose curated install shapes without creating a second safety model.
 - Checked-in example fixtures now cover financial UI, CI hardening, and quality lifecycle evidence flows.
-- A third-party case study records a NEKOWORK run against `sindresorhus/is-plain-obj`.
-- Public npm alpha metadata is prepared for `0.1.0-alpha.0`; publish execution remains blocked on npm owner auth.
+- Third-party case studies record NEKOWORK runs against `sindresorhus/is-plain-obj`, `jshttp/basic-auth`, and `python-hyper/h11`.
+- Public npm alpha `0.1.0-alpha.2` is published under the `alpha` dist-tag.
 
 ## Remaining Optional Work
 
 | Item | Priority | Reason |
 |---|---|---|
-| Public npm publish execution | High | Requires npm owner login and 2FA readiness |
-| More third-party case studies | Medium | One public repo case study exists; more languages/frameworks would improve adoption evidence |
-| Internal provider adapter | Low until requested | Only useful for private infrastructure |
-| More skill catalog expansion | Low | Should stay selective to preserve progressive disclosure |
+| Stable `latest` promotion | Medium | `alpha` is correct; npm keeps `latest` on the first alpha line for now, so move it to a stable version later |
+| More third-party case studies | Low | Three public repo case studies exist; more frameworks can still improve adoption evidence later |
+| More skill catalog expansion | Low | Catalog expansion should stay selective to preserve progressive disclosure |
 
 ## Explicit Non-Goals
 
-- No public npm publish for `0.0.3`; public alpha publish requires npm owner auth.
+- No public npm publish for `0.0.3`; public alpha starts at `0.1.0-alpha.0`.
 - No automatic promotion of learned instincts without human confirmation.
 - No tmux-first runtime import from OMC.
 - No bulk import of large external skill catalogs.
@@ -105,10 +113,10 @@ Current local result for this working tree:
 
 ## External Readiness Score
 
-Current external readiness, excluding npm publish execution and broader adoption evidence: **8.8 / 10**.
+Current external readiness, excluding broader adoption evidence: **9.1 / 10**.
 
 Main deductions:
 
-- No public npm package yet because npm owner auth is not active on this machine.
-- Only one independent real-world external project case study so far.
+- `latest` currently remains on the first alpha; docs still recommend `@alpha` until a stable release exists.
+- Three independent real-world external project case studies exist so far.
 - Advanced surfaces exist but are intentionally secondary to the public decomposed workflow and install flow.

package/docs/CATALOG-PACKS.md

@@ -0,0 +1,77 @@
+# Catalog Packs
+
+NEKOWORK intentionally keeps the catalog selective. Every agent, skill, hook, module, profile, and pack must preserve the verification loop:
+
+```text
+Claude work -> Codex verification -> report -> Human Gate -> explicit apply
+```
+
+Packs are public install aliases over validated profiles. They make the catalog easier to choose without creating a second safety model.
+
+## Current Shape
+
+```text
+7 official packs
+9 install profiles
+7 modules
+36 components
+11 agents
+10 skills
+5 hooks
+5 harness targets
+6 case-study flows
+245 tests
+```
+
+Harness targets:
+
+```text
+Claude, Codex, Cursor, Gemini, OpenCode
+```
+
+Case-study flows:
+
+```text
+financial UI mock
+GitHub Actions hardening
+quality lifecycle smoke
+npm package boundary
+auth parser boundary
+Python protocol parser boundary
+```
+
+## Official Packs
+
+| Pack | Profile | Best For | Representative Workflow |
+|---|---|---|---|
+| `core` | `core` | Minimal verification runtime | `doctor -> ask -> run -> report -> gate` |
+| `quality` | `quality` | Disciplined development and evidence coverage | `ask --profile quality -> run --profile quality --strict-quality -> report` |
+| `security` | `security` | Auth, secrets, permissions, deploy, financial, or data-sensitive changes | `ask --profile security -> run --profile security --secure --strict-quality -> report -> gate` |
+| `frontend` | `frontend` | UI mockups, component review, accessibility-oriented checks | `ask --profile product -> team -> run -> report` |
+| `testing` | `testing` | Regression planning and coverage-oriented handoffs | `plan -> work -> verify --profile quality --strict-quality -> report` |
+| `release` | `developer` | Release readiness, changelog, and no-ship/ship evidence | `run -> report -> gate -> ship` |
+| `enterprise` | `full` | Full stable catalog evaluation with all gates intact | `ask -> plan -> team -> work -> verify -> gate -> ship -> report -> apply` |
+
+## Commands
+
+```bash
+node scripts/install-plan.js --list
+node scripts/install-plan.js --pack security
+node scripts/install-plan.js --pack quality --target claude --json
+node scripts/install-apply.js --pack core --project-root <target>
+```
+
+`--pack` and `--profile` cannot be used together. A pack resolves to exactly one profile, and profile safety validation still rejects any default that weakens Codex verification, Human Gate, or single-executor mutation policy.
+
+## Positioning
+
+NEKOWORK does not try to be the largest catalog. It is a curated catalog for a reportable evidence pipeline:
+
+```text
+selective catalog
++ multi-surface projection
++ evidence report
++ Human Gate
++ explicit apply
+= local-first AI development quality runtime
+```