用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
直接命令不会经过审查 Prompt；运行前请先检查来源。
npx skills add https://github.com/InugamiDev/ultrathink-oss --skill plan-archive命令会保持在同一行。复制前请横向滚动并检查完整内容。
想先保存到本地？可下载 SkillsMP 当前能够提供的文件。
正在显示 SKILL.md
| name | plan-archive |
| description | Archive completed plans, extract lessons learned, and write journey journals to long-term memory |
| layer | hub |
| category | workflow |
| triggers | ["/plan-archive","archive this plan","plan is done","wrap up the plan","write a retrospective"] |
| inputs | [{"plan":"The completed (or abandoned) implementation plan"},{"outcome":"What actually happened -- successes, deviations, failures (optional, will be inferred from context)"},{"journalNotes":"User-provided reflections or notes (optional)"}] |
| outputs | [{"archive":"Structured archive entry with plan summary, outcome, and metadata"},{"journeyJournal":"Narrative lessons-learned document for long-term memory"},{"patterns":"Extracted reusable patterns or anti-patterns"}] |
| linksTo | ["kanban","research"] |
| linkedFrom | ["plan","cook","team","ship"] |
| preferredNextSkills | ["kanban","brainstorm"] |
| fallbackSkills | ["plan"] |
| riskLevel | low |
| memoryReadPolicy | selective |
| memoryWritePolicy | selective |
| sideEffects | ["Writes archive entry to memory","Writes journey journal to memory","Updates kanban board status if applicable"] |
Close the loop on completed or abandoned plans by extracting what happened, why it happened, and what future plans should learn from it. This skill transforms lived experience into institutional memory.
Without archiving, every plan starts from zero. With archiving, each plan benefits from all prior plans.
Structure:
# Journey Journal: [Plan Title]
## The Mission
[What we set out to do, in plain language]
## What Actually Happened
[Narrative of the execution -- the key turning points, surprises, and decisions]
## What We Learned
[Bullet-pointed lessons, each with enough context to be useful standalone]
## What We Would Do Differently
[Specific, actionable changes for next time]
## Artifacts Worth Keeping
[Code patterns, configs, approaches that should be remembered]
If this archive is for a GSD project (.planning/ exists with SPEC.md):
Run the verification gate before archiving:
source ~/.claude/hooks/../.claude/hooks/gsd-utils.sh # or the UltraThink hooks dir
# Equivalent logic: check SPEC → PLAN traceability + VERIFICATION.md completeness
Check traceability: Every SPEC.md acceptance criterion MUST have a corresponding must-have in at least one PLAN.md
Check verification: Every must-have in VERIFICATION.md must be ✅ PASS — no ❌ FAIL or STUB remaining
If gates fail: Do NOT archive. Report gaps to user. Create fix plans if needed.
If gates pass: Proceed to archive.
This prevents incomplete work from being archived as "done".
learn-pattern skill. Successes become offensive adaptations; failures turn the wheel./tmp/ultrathink-progress-* bridge files.# Archive: [Plan Title]
## Metadata
- **Created**: [date]
- **Archived**: [date]
- **Status**: Completed | Completed with Deviations | Partial | Abandoned
- **Phases completed**: [N] of [M]
- **Overall complexity**: [as experienced, not as planned]
## Plan Summary
[2-3 sentence summary of what the plan aimed to do]
## Outcome Summary
[2-3 sentence summary of what actually happened]
## Phase Outcomes
| Phase | Planned | Actual | Complexity (Plan→Actual) |
|-------|---------|--------|--------------------------|
| 1: [name] | [goal] | [outcome] | Low → Medium |
| 2: [name] | [goal] | [outcome] | Medium → Medium |
## Assumption Audit
| Assumption | Expected | Actual |
|-----------|----------|--------|
| [A1] | True | True |
| [A2] | True | False — [what was actually true] |
## Risk Audit
| Risk | Predicted | Materialized? | Notes |
|------|-----------|---------------|-------|
| [R1] | Medium likelihood | No | |
| [R2] | Low likelihood | Yes | [what happened] |
## Lessons Learned
1. **[lesson title]**: [explanation]
2. **[lesson title]**: [explanation]
## Patterns Extracted
- **Pattern**: [name] — [description and when to apply]
- **Anti-pattern**: [name] — [description and what to do instead]
## Tags
[searchable tags for future retrieval: technology, domain, problem-type]
/plan-archive
/plan-archive The WebSocket migration took twice as long as expected because we underestimated the testing surface. The fallback polling mechanism we added was not in the original plan but turned out to be essential.
/plan-archive We abandoned the GraphQL migration because the team decided to use tRPC instead. Key learning: evaluate alternatives before committing to a migration path.
Plan: "Add dark mode to the app" Status: Completed with deviations Key finding: Phase 2 (component migration) was estimated as Medium but was actually High because 12 third-party components did not respect CSS variables. Lesson: "Always audit third-party component theming support before estimating a theme system migration." Pattern: "Create a compatibility matrix for third-party components before starting theme work."
Plan: "Migrate from REST to GraphQL" Status: Abandoned at Phase 1 Key finding: Schema definition revealed that the existing data model was not graph-shaped; many endpoints were simple CRUD. Lesson: "Match API paradigm to data shape. GraphQL is not always better than REST." Anti-pattern: "Adopting a technology because it is popular rather than because it fits the problem."
Unified design foundations — design system architecture, tokens, component specs, visual principles, creative vision, figma integration, plus brand design system loader (66 real brands via DESIGN.md). Absorbs design, design-system, design-systems, design-principles, design-router, creative-vision, figma, design-md.
Render, summarize, and present markdown documents and structured content in multiple output modes
Ultra UI skill - combines Google's DESIGN.md spec (machine-readable design tokens) with the ui-ux-pro-max knowledge base (91 styles, 161 palettes, 73 font pairings, 161 products, 104 UX guidelines, 25 chart types). Generates lint-clean DESIGN.md files, validates token references and WCAG contrast, exports Tailwind/DTCG tokens, and diffs design systems version-over-version.
基于 SOC 职业分类
