Releases: guglxni/chronos
Releases · guglxni/chronos
v2.0.0 — chronos-data
chronos-data 2.0.0
Autonomous Data Incident Root Cause Analysis Agent — now published to PyPI.
pip install chronos-dataWhat CHRONOS does
When a data quality test fails, CHRONOS runs a 10-step autonomous investigation — walking OpenMetadata lineage, querying a Graphiti temporal knowledge graph, checking code commits — and returns a structured incident report with root cause, confidence score, evidence chain, and recommended actions. No engineer required.
What's in this release
- 10-step LangGraph RCA pipeline — entity resolution → test history → temporal changes → schema diff → upstream lineage → code search → downstream impact → audit events → LLM synthesis → PROV-O provenance
- Native MCP server (stdio + SSE) with 7 tools:
trigger_investigation,get_incident,list_incidents,query_lineage,search_entity,get_graph_context,poll_failures - A2A agent discovery card at
/.well-known/agent-card.json— self-configuring for OpenClaw, Hermes, AutoGen, Claude - 24/7 autonomous monitoring — polls OpenMetadata continuously, deduplicates, auto-triggers investigations
- W3C PROV-O provenance — JSON-LD / Turtle / PROV-N audit artifacts per investigation
- React frontend on Vercel with live demo, dashboard, investigation detail, and For Agents section
- Agent skill published to 39 agents:
npx skills add guglxni/chronos-skill
Install
# Python package
pip install chronos-data
# Agent skill (Claude Code, OpenClaw, Cursor, Hermes, +35 more)
npx skills add guglxni/chronos-skill
# MCP server
chronos-mcp # stdio (Claude Desktop)
chronos-mcp --transport sse --port 8101 # SSE (remote agents)Links
- 🌐 Live demo: https://chronos-frontend-eight.vercel.app
- 📡 API: https://chronos-api-0e8635fe890d.herokuapp.com
- 📖 API docs: https://chronos-api-0e8635fe890d.herokuapp.com/docs
- 🤖 Agent card: https://chronos-api-0e8635fe890d.herokuapp.com/.well-known/agent-card.json
- ⚡ Agent skill: https://github.com/guglxni/chronos-skill
- 📦 PyPI: https://pypi.org/project/chronos-data/2.0.0/
v0.0.1 — AIDLC Inception Phase Complete
⏳ CHRONOS — Inception Phase Complete
What's included
AIDLC Workflow Setup (AI-DLC v0.1.7)
- Core workflow rules (AGENTS.md)
- Full rule details (.aidlc-rule-details/)
- Workflow state tracking (aidlc-docs/aidlc-state.md)
- Audit trail (aidlc-docs/audit.md)
Inception Phase Artifacts
- 📋 Requirements Analysis — 10 functional + 5 non-functional requirements (comprehensive depth)
- 👤 User Stories — 3 personas (Priya, Alex, Meera), 6 epics, 16 stories with acceptance criteria
- 🗺️ Workflow Plan — 6 construction units with dependency ordering
- 🏗️ Application Design — Triple-graph architecture, component inventory, data flow diagrams
- 📦 Units Specification — File-level component inventory for all 6 units
Project Foundation
- README.md with architecture diagram, tech stack, quick start guide
- .gitignore, .env.example, LICENSE (Apache 2.0)
- GitHub: 20 topics, 14 custom labels, milestone, 8 issues
Next Steps
Construction Phase begins with Unit 1: Core Infrastructure & Configuration
Tech Stack
OpenMetadata | Graphiti | GitNexus | LangGraph | LiteLLM | FastAPI | React + React Flow | Docker Compose