菜单
Trigger self-learning synthesis on demand. Analyzes recent session learnings, clusters patterns, and updates MEMORY.md with verified insights. Use when user says /self-learn, wants to consolidate learnings, or needs to manually trigger the synthesis that normally runs on session end.
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。
基于 SOC 职业分类
| name | self-learn |
| description | Trigger self-learning synthesis on demand. Analyzes recent session learnings, clusters patterns, and updates MEMORY.md with verified insights. Use when user says /self-learn, wants to consolidate learnings, or needs to manually trigger the synthesis that normally runs on session end. |
| user-invocable | true |
| arguments | [{"name":"force","description":"Run even if below the 5-learning threshold","required":false}] |
Trigger the self-learning synthesis to analyze recent session learnings and update MEMORY.md.
learnings.db (patterns, gotchas, decisions, tool errors, pivots)--force)/self-learn # Normal run (respects 5-learning threshold)
/self-learn --force # Force run even below threshold
Run this command:
bash /Users/quartershots/Source/.claude/scripts/self-learn.sh $ARGS
Where $ARGS is --force if the user specified force.
A LaunchAgent runs this every 12 hours automatically (threshold-gated):
~/Library/LaunchAgents/com.claude.self-learn.plistlaunchctl load ~/Library/LaunchAgents/com.claude.self-learn.plistlaunchctl unload ~/Library/LaunchAgents/com.claude.self-learn.plistlaunchctl list | grep self-learnUpdates /Users/quartershots/Source/.claude/projects/-Users-quartershots-Source/memory/MEMORY.md with:
Design consultation: understands your product, researches competitors, proposes a complete design system (aesthetic, typography, color, layout, spacing, motion), and generates font+color preview pages. Creates DESIGN.md as your project's design source of truth. Use when asked to "design my app", "create a design system", "pick fonts and colors", "design consultation", or "make it look good". For existing sites, use /plan-design-review to infer the system instead.
Post-ship documentation update. Reads all project docs, cross-references the diff, updates README/ARCHITECTURE/CONTRIBUTING/CLAUDE.md to match what shipped, polishes CHANGELOG voice, cleans up TODOS, and optionally bumps VERSION. Use when asked to "update docs", "sync documentation", "document what shipped", "update the readme after shipping", or "post-release docs".
Eng manager-mode plan review. Lock in the execution plan — architecture, data flow, diagrams, edge cases, test coverage, performance. Walks through issues interactively with opinionated recommendations. Use when asked to "review my plan", "eng review", "check my architecture", "review before I build", or "sanity check this plan".
Pre-landing PR review. Analyzes diff against the base branch for SQL safety, LLM trust boundary violations, conditional side effects, and other structural issues. Use when asked to "review my code", "review this PR", "check before merge", "pre-landing review", or "look at my diff".
Ship workflow: detect + merge base branch, run tests, review diff, bump VERSION, update CHANGELOG, commit, push, create PR. Use when asked to "ship it", "ship this branch", "merge and deploy", "create a PR", "land this", or "push to production".
Fast headless browser for QA testing and site dogfooding. Navigate any URL, interact with elements, verify page state, diff before/after actions, take annotated screenshots, check responsive layouts, test forms and uploads, handle dialogs, and assert element states. ~100ms per command. Use when you need to test a feature, verify a deployment, dogfood a user flow, or file a bug with evidence.