Darkplex Intelligence Layer v0.2.0 — governance, knowledge extraction, self-improvement
|
Some checks failed
Tests / test (push) Failing after 5s
- Renamed to 'Preference & Learning Pair Extractor' - NEW: Positive reinforcement detection (praise, affirmation, emoji, accept+continue) - NEW: Teaching moment detection (rules, explanations, reminders, preferences) - NEW: Soft redirect detection (let's rather, alternative plan, switch to) - Outputs: DPO pairs, SFT pairs (alpaca format), teaching pairs - Improved false positive filters (subagent output, apt, system messages) - 4x more training signal: 3 → 11 pairs from same 30-day window - 446 tests passing |
||
|---|---|---|
| .forgejo/workflows | ||
| cortex | ||
| tests | ||
| .gitignore | ||
| Dockerfile | ||
| pyproject.toml | ||
| README.md | ||
Cortex 🧠
Intelligence layer for OpenClaw. The higher-order thinking above the nervous system.
Modules
| Module | Command | What it does |
|---|---|---|
| Triage | cortex triage score "text" |
Priority scoring (urgency × importance × effort × readiness) |
| Health Scanner | cortex health [--json] |
Proactive system health checks |
| Feedback Loop | cortex feedback --since 6h |
Extract lessons from session transcripts |
| Memory Hygiene | cortex hygiene stats|stale|duplicates|orphans|archive |
Keep memory clean |
| Roadmap | cortex roadmap list|add|overdue|report |
Task tracking with deadlines & dependencies |
| Validate | cortex validate --transcript <path> |
Sub-agent output validation |
| Enhanced Search | cortex search "query" |
Multi-strategy memory search |
| Auto Handoff | cortex handoff <file> |
Generate session handoff documents |
Install
pip install -e .
Configuration
Cortex uses environment variables for path configuration. All paths have sensible defaults (~/.cortex/).
| Variable | Default | Description |
|---|---|---|
CORTEX_HOME |
~/.cortex |
Base directory |
CORTEX_MEMORY_DIR |
$CORTEX_HOME/memory |
Memory files |
CORTEX_CONFIG |
$CORTEX_HOME/config.json |
Config file |
CORTEX_GROWTH_LOG |
$CORTEX_MEMORY_DIR/growth-log.md |
Feedback loop output |
CORTEX_ROADMAP |
$CORTEX_MEMORY_DIR/roadmap.json |
Roadmap data |
Optional config.json for customization:
{
"permanent_files": ["MEMORY.md", "WORKING.md", "README.md"],
"sessions_dir": "~/.openclaw/agents/main/sessions"
}
Docker
docker build -t cortex .
docker run cortex triage score "production down"
Tests
python -m pytest tests/ -v
Architecture
Cortex is a standalone Python package that OpenClaw calls via CLI. No tight coupling, no plugin system — just clean Unix-style commands.
OpenClaw (Node/TS)
│
├── cortex triage score "..." ← CLI calls
├── cortex health --json
├── cortex feedback --since 6h
└── cortex hygiene stats
│
└── Cortex (Python)
├── triage.py
├── health_scanner.py
├── feedback_loop.py
├── memory_hygiene.py
├── roadmap.py
├── validate_output.py
├── enhanced_search.py
└── auto_handoff.py
License
MIT