Skip to content

CoMatrixlabs/asteraskills

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

10 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

asteraskills

Open-source security intelligence tooling: LangChain StructuredTools (CVE, EPSS, CISA KEV, ATT&CK, CWE/CAPEC, exploits, compliance frameworks, OTX, and more), a LangGraph orchestrator for natural-language questions, and an optional Risk Scanner CLI for multi-domain artifact analysis (SBOM, code, IaC, logs, evidence).

Licensed under Apache-2.0 (see LICENSE).

Requirements

  • Python 3.10+
  • Optional: PostgreSQL and/or Qdrant (or Chroma) for semantic search and indexed intel; many API-based tools work with only network access and API keys.

Install

git clone <your-fork-or-upstream-url>
cd asteraskills
python -m venv .venv
source .venv/bin/activate   # Windows: .venv\Scripts\activate
pip install -e ".[runtime]"

Copy .env.example to .env and set at least the LLM and embedding keys your setup needs (see docs/bring-your-own-llm.md).

CLI

Command Purpose
asteraskills doctor Verify imports and TOOL_REGISTRY load
asteraskills list List registered tools and descriptions
asteraskills run <tool> --args '<json>' Run one tool; JSON on stdout
asteraskills agent run "<question>" LangGraph CVE/CWE/ATT&CK-style orchestration
risk-scanner Scan artifacts (see risk-scanner --help)

Example:

asteraskills run cve_details --args '{"cve_id":"CVE-2024-3400"}'

Repository layout

Path Role
app/asteraskills/ Core package: tools/, agents/, ingestion/, storage/, config/
risk_scanner/ Standalone scanner CLI, analyzers, reporters (SARIF, JSON, Markdown, CSV, HTML)
skills/ Claude Agent Skill definitions (SKILL.md per skill)
tests/ Pytest
docs/ Architecture and LLM configuration

Documentation

  • Setup, usage, and data access — purpose, install tiers, CLI examples, phased rollout (direct DB for early adopters → hosted API with tokens and rate limits).
  • Architecture — how the tool registry, agents, storage, and Risk Scanner fit together.
  • Bring your own LLM — OpenAI vs Anthropic, models, embeddings, optional config/llm_models.yaml.
  • Enrichment chain — Risk Scanner enrichment pipeline (ATT&CK and controls).

Claude Code / Agent Skills

Point your agent at the skill folders under skills/ (for example skills/astera-security-intel/). Each skill’s SKILL.md describes prompts and when to use the bundled tools.

Contributing

Issues and pull requests are welcome. Use ruff for style (pip install -e ".[dev]" then ruff check app risk_scanner tests).

About

No description, website, or topics provided.

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors

Languages