Menu
Efficient context state inspection, task lifecycle management, and session tracking
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
Automated visual testing with Playwright MCP - test web apps, presentations, websites, and documents with scalable reviewer perspectives
Expert guidance on RAG (Retrieval-Augmented Generation) system design including chunking strategies, embedding selection, retrieval methods, and vector database choices
Design complex system architectures, evaluate tradeoffs, and make critical technical decisions requiring deep reasoning
Deep code analysis identifying subtle bugs, security issues, performance problems, and architectural concerns requiring expert-level reasoning
Systematic codebase onboarding. Builds a mental model of a new or unfamiliar project by exploring structure, architecture, key data models, entry points, and auth patterns.
Systematically extract knowledge from scattered documentation to prevent catastrophic forgetting. Creates structured extraction reports with status tags.
| name | context |
| description | Efficient context state inspection, task lifecycle management, and session tracking |
| short_desc | /context state inspection and session tracking |
| keywords | ["context state","CONTEXT_STATE","session tracking","task lifecycle","context size","/context","what's my context","context usage","current task state"] |
| model | haiku |
Purpose: Efficient context state inspection, task lifecycle management, and session tracking.
Model: Haiku 4.5 (sub-second execution, token-efficient getter commands)
Token Cost: VERY LOW (getter commands, not full file reads)
Claude Code should proactively suggest /context commands when:
IF (user asks "what's the status?") → suggest /context status
IF (user seems lost) → suggest /context summary
IF (Claude context filled >60%) → suggest /context summary
IF (session >30 minutes) → suggest /context update
IF (switching tasks) → suggest /context pause
IF (completed milestone) → suggest /context complete
/context status
Returns single-line task status (~50 tokens).
/context summary
Returns status + current work + blockers (~500 tokens).
/context blockers
Returns just current blockers (~100 tokens).
/context log [n]
Returns last N log entries (~150 tokens per 5 entries).
/context size
Returns line count, token estimate, warns if >200 lines.
/context pause [optional-note]
Saves to TEMP_MEMORY_[task].md, resets to IDLE.
/context resume [taskname]
Restores from TEMP_MEMORY_[task].md.
/context complete [optional-note]
Archives to .claude/context/archive/, resets to IDLE.
/context update "Message"
Adds timestamped entry to Session Log.
Before (reading full CONTEXT_STATE.md):
After (using /context commands):
/context update after every 5-10 actions/context pause when switching tasks/context status silently before expensive operations