system-overview
Scannednpx machina-cli add skill parcadei/Continuous-Claude-v3/system_overview --openclawFiles (1)
SKILL.md
2.7 KB
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
| Component | Location |
|---|---|
| Hook registration | .claude/settings.json |
| Hook implementations | .claude/hooks/src/*.ts |
| Rules (auto-injected) | .claude/rules/*.md |
| Skills | .claude/skills/*/SKILL.md |
| Setup wizard | opc/scripts/setup/wizard.py |
| Recall script | opc/scripts/recall_temporal_facts.py |
| Store learning | opc/scripts/core/store_learning.py |
| Symbol index builder | opc/scripts/build_symbol_index.py |
Environment Variables
| Variable | Purpose |
|---|---|
CONTINUOUS_CLAUDE_DB_URL | PostgreSQL connection |
VOYAGE_API_KEY | Embeddings (optional) |
BRAINTRUST_API_KEY | Tracing (optional) |
CLAUDE_PROJECT_DIR | Auto-set by Claude Code |
Source
git clone https://github.com/parcadei/Continuous-Claude-v3/blob/main/.claude/skills/system_overview/SKILL.mdView on GitHub Overview
This skill explains how the opinionated Continuous Claude setup works, focusing on the Memory Layer, Hooks, and Coordination. It highlights how past-session facts are stored and queried, how event-driven hooks manage context, and how handoffs coordinate subagents for continuity.
How This Skill Works
Technically, the system uses a Memory Layer (PostgreSQL + pgvector) to store 78,000+ temporal facts with semantic search via embeddings. Hooks (nine event types) trigger actions at key points like SessionStart, UserPromptSubmit, and SessionEnd to manage continuity, context injection, and state saving. The Continuity System records learnings and creates handoffs through a ledger and coordinated subagents, with recall scripts and a TLDR code intelligence pipeline aiding fast topic extraction.
When to Use It
- User asks "how does this work?"
- User asks "what can you remember?"
- User asks "what's different about this setup?"
- User runs `/system_overview`
- You want to verify or explain the setup options (SQLite vs PostgreSQL + pgvector) and environment variables
Quick Start
- Step 1: Run the setup wizard to configure storage: uv run python opc/scripts/setup/wizard.py and choose PostgreSQL + pgvector (semantic search).
- Step 2: Issue /system_overview to display the architecture and components (Memory Layer, Hooks, Continuity System).
- Step 3: Review the Key Files and Environment Variables (e.g., CONTINUOUS_CLAUDE_DB_URL, embeddings API key) to tailor the setup.
Best Practices
- Refer to Memory Layer details: 78,000+ temporal facts, automatic learnings, and semantic search with embeddings to illustrate retention.
- Highlight the nine Hook types (SessionStart, UserPromptSubmit, PreToolUse, PostToolUse, PreCompact, Stop, SubagentStart, SubagentStop, SessionEnd) and their roles in context management.
- Explain the Continuity System structure: the ledger path and handoffs location to show how persistence and transfer work.
- Point to the recall workflow (recall_temporal_facts.py) for retrieving relevant past facts.
- Recommend running the setup wizard to choose storage options (SQLite vs PostgreSQL + pgvector) and review environment variables like CONTINUOUS_CLAUDE_DB_URL
Example Use Cases
- A user asks how this works and runs /system_overview to see the high-level architecture.
- An engineer wants to know what the system can remember and checks the Memory Layer details (temporal facts and embeddings).
- A maintainer asks what's different about this setup and reviews the Hooks and Continuity System sections.
- During a new session, the system loads the continuity ledger via SessionStart and rebuilds the symbol index.
- A developer validates recall capabilities by running recall_temporal_facts.py to fetch topic-specific memories.
Frequently Asked Questions
Add this skill to your agents
