Menu
Record final outcomes to history.md, not intermediate requests or reversed decisions
Installer avec Codex ou Claude Copiez ce prompt, collez-le dans Codex, Claude ou un autre assistant, puis laissez-le vérifier la page du skill et l'installer pour vous.
Basé sur la classification professionnelle SOC
{what this skill teaches agents}
{what this skill teaches agents}
**WORKFLOW SKILL** — Iteratively improve skill frontmatter compliance using the Ralph loop pattern. WHEN: "run sensei", "sensei help", "improve skill", "fix frontmatter", "skill compliance", "frontmatter audit", "score skill", "check skill tokens". INVOKES: token counting tools, test runners, git commands. FOR SINGLE OPERATIONS: use token CLI directly for counts/checks.
**WORKFLOW SKILL** — Research, outline, and scaffold a Slidev presentation. Phase 1 researches the topic and generates a detailed OUTLINE.md via runSubagent research and user guidance. Phase 2 generates slides.md, style.css, and images/ from the approved outline. WHEN: "create presentation", "new presentation", "scaffold presentation", "new Slidev deck", "new talk", "create slides", "start a presentation", "presentation about". INVOKES: runSubagent (web search, Microsoft Learn, WorkIQ), vscode_askQuestions, file-system tools. FOR SINGLE OPERATIONS: copy template folder manually.
**WORKFLOW SKILL** — Query WorkIQ for recent activity signals, identify patterns in conversations and meetings, cross-reference with latest tech announcements, and propose aligned demos as GitHub issues. WHEN: "intelligence scan", "what demos to build", "scan recent activity", "WorkIQ scan", "demo proposals from conversations", "identify demo opportunities", "what should I demo next". INVOKES: WorkIQ MCP, web search, GitHub issue tools. FOR SINGLE OPERATIONS: Query WorkIQ MCP directly.
Build an n-layer dependency graph from .NET upgrade assessment data and generate a phased modernization plan. Parses assessment.json from the .NET upgrade tools, extracts project dependency hierarchy via breadth-first layering, splits thick layers to cap projects per sub-layer, and produces a phased upgrade plan with Mermaid diagrams. Each phase targets PR-reviewable scope. SDK-style conversion is Phase 0, then layer-by-layer bottom-up modernization. WHEN: "plan appmod phases", "generate upgrade layers", "dependency layer plan", "phased migration plan", "break down dotnet upgrade", "layer extraction", "modernization phases", "upgrade dependency graph", "PR-sized migration plan", "appmod phasing strategy".
| name | history-hygiene |
| description | Record final outcomes to history.md, not intermediate requests or reversed decisions |
| domain | documentation, team-collaboration |
| confidence | high |
| source | earned (Kobayashi v0.6.0 incident, team intervention) |
History files (.md files tracking decisions, spawns, outcomes) are read cold by future agents. Stale or incorrect entries poison decision-making downstream. The Kobayashi incident proved this: history said "Brady decided v0.6.0" when Brady had reversed that to v0.8.17. Future spawns read the wrong truth and repeated the mistake.
✓ Correct:
✗ Incorrect:
