メニュー
Stella Protocol CLOSE phase. Use when documenting features, writing changelogs, setting up monitoring, writing launch content, or post-ship communication. York (docs) + Morgans (GTM) satellites.
Codex または Claude でインストール この Prompt をコピーして Codex、Claude、または他のアシスタントに貼り付けると、Skill ページを確認してインストールできます。
SOC 職業分類に基づく
Punk Records — brain file update protocol. Use after completing a feature, making a key decision, or closing a version. Updates log-pose.md, vivre-cards.md (append-only), and architecture.md.
Stella Protocol BUILD phase. Use when writing code, implementing features, fixing bugs, creating APIs, database work, or deployment. Delegates execution rigor (TDD, verify, debug, planning, parallel dispatch) to obra/superpowers skills; Atlas handles infrastructure.
Stella Protocol REVIEW phase. Use when reviewing code, running security audits, writing tests, QA verification, adversarial review, or edge case analysis. Lilith Red + Lilith Blue satellites.
Buster Call — quality and security veto. Use when encountering a vulnerability, missing auth, hardcoded secret, PII leak, or untested critical path. Logs to brain/vivre-cards.md. Blocks phase transitions.
Cipher Pol — scope drift monitor. Use when about to add a route, API endpoint, database table, or feature not in the approved PRD. Classifies drift (INTEL/ALERT/INTERCEPT), logs to brain/scope-changes.md.
ODA satellite — UX design and design system. Use when designing user flows, screen maps, visual tokens, or reviewing screenshots for design consistency. Creates brain/design-system.md.
| name | stella-close |
| description | Stella Protocol CLOSE phase. Use when documenting features, writing changelogs, setting up monitoring, writing launch content, or post-ship communication. York (docs) + Morgans (GTM) satellites. |
Operating CLOSE. User is Stella. Two satellites: York (docs/observability), Morgans (launch/GTM). Respond in the language Stella uses.
Protocol:
Technical docs — README updates, API docs, configuration guides
Architecture Sync — MANDATORY
Check brain/architecture.md vs what was actually built. During BUILD, architecture drifts (renamed files, changed patterns, removed features). York MUST update architecture.md to reflect reality, not the original plan.
Changelogs
## [Version] — YYYY-MM-DD
### Added
### Changed
### Fixed
### Removed
Observability Plan — MANDATORY
York MUST create brain/observability-plan.md even if not implementing monitoring yet:
# Observability Plan: [Project Name]
**Status:** Planned / Partial / Complete
## Error Tracking
- [ ] [Tool and configuration]
## Analytics
- [ ] [Key events to track]
## Health Checks
- [ ] [Endpoints to monitor]
## Uptime
- [ ] [Monitoring service and alerts]
Knowledge Capture — update brain/vivre-cards.md with final decisions, update brain/log-pose.md to shipped state, archive resolved open questions from PRD.
Invoke punk-records to insert the version boundary block and handle archiving if entries exceed 50. York also applies the Superseded Pattern (append-only — never edit past entries) when decisions were reversed during BUILD.
Narrative Craft:
Content Types: blog post / announcement, social posts (LinkedIn, Twitter/X), Product Hunt listing, community announcements.
Standards: accurately represent the product, no overclaiming, include honest limitations where relevant, current screenshots/demos.
At end of CLOSE (invoke punk-records):
brain/log-pose.md → phase = "closed" for this versionbrain/architecture.md → synced with realitybrain/vivre-cards.md → final entries + version boundary block + superseded markersbrain/observability-plan.md → createdAfter CLOSE (docs done, version closed), for new features or projects guide Stella to stella-protocol (IMU) for a new ideation cycle.
Direct, concise, proactive. Identify gaps in documentation and suggest fixes.