L’interface est localisée ; les présentations et documents des Skills sont traduits automatiquement par IA dans votre langue. Le code, les commandes, chemins et termes techniques restent en version originale, et le document de l’auteur fait foi.
Explorer les Skills
Productivité · SKILL.md

system-overview

system-overview

Show users how Continuous Claude works - the opinionated setup with hooks, memory, and coordination

Independent GitHub snapshot · Automated classification · Original documentation governs

Documentation de la Skill

GitHub

name: system-overview description: Show users how Continuous Claude works - the opinionated setup with hooks, memory, and coordination

System Overview

Show users how Continuous Claude works - the opinionated setup with hooks, memory, and coordination.

When to Use

  • User asks "how does this work?"
  • User asks "what can you remember?"
  • User asks "what's different about this setup?"
  • User runs /system_overview

Response

CONTINUOUS CLAUDE SYSTEM OVERVIEW
=================================

MEMORY LAYER (PostgreSQL + pgvector)
------------------------------------
- 78,000+ temporal facts from past sessions
- Learnings extracted automatically at session end
- Semantic search with embeddings

RECALL: uv run python opc/scripts/recall_temporal_facts.py --query "your topic"

HOOKS (9 event types registered)
--------------------------------
SessionStart    → Load continuity ledger, rebuild symbol index
UserPromptSubmit → Skill activation check, context injection
PreToolUse      → Smart search routing (Grep → TLDR for code)
PostToolUse     → File claims, compiler feedback
PreCompact      → Save state before context compaction
Stop            → Extract learnings, create handoffs
SubagentStart   → Register spawned agents
SubagentStop    → Coordination, handoff creation
SessionEnd      → Cleanup

CONTINUITY SYSTEM
-----------------
Ledger:   thoughts/ledgers/CONTINUITY_CLAUDE-{session}.md
Handoffs: thoughts/shared/handoffs/{session}/*.yaml

Commands:
  /resume_handoff <path>  - Continue from handoff
  /create_handoff         - Create snapshot for transfer

TLDR CODE INTELLIGENCE
----------------------
5-layer analysis: AST → Call Graph → CFG → DFG → PDG
95% token savings vs reading raw files
Auto-intercepts Grep for .py/.ts/.go/.rs files

Pre-built index: /tmp/claude-symbol-index/symbols.json

SETUP
-----
Run: uv run python opc/scripts/setup/wizard.py

Options:
  [1] SQLite only (simple, offline)
  [2] PostgreSQL + pgvector (semantic search)

Key Files

ComponentLocation
Hook registration.claude/settings.json
Hook implementations.claude/hooks/src/*.ts
Rules (auto-injected).claude/rules/*.md
Skills.claude/skills/*/SKILL.md
Setup wizardopc/scripts/setup/wizard.py
Recall scriptopc/scripts/recall_temporal_facts.py
Store learningopc/scripts/core/store_learning.py
Symbol index builderopc/scripts/build_symbol_index.py

Environment Variables

VariablePurpose
CONTINUOUS_CLAUDE_DB_URLPostgreSQL connection
VOYAGE_API_KEYEmbeddings (optional)
BRAINTRUST_API_KEYTracing (optional)
CLAUDE_PROJECT_DIRAuto-set by Claude Code
Skills enregistrées