Skip to content

aveproject/ave-api

ave-api

Read-only reference implementation of the AVE (Agentic Vulnerability Enumeration) standard's record lookup API.

OpenSSF Scorecard CodeQL

What this is, and isn't

This is a reference implementation, not the standard. AVE itself, the records and schema, lives in aveproject/ave. This service is a thin, read-only lookup layer over that data, provided so implementers have a working example to call or fork rather than parsing raw JSON files themselves. Anyone can run their own instance; nothing about AVE depends on this specific deployment staying available.

Endpoints

Method Path Description
GET / Service info, record count, cache age
GET /records All records
GET /records/{ave_id} One record
GET /records/{ave_id}/mitigation The record's neutral mitigation object only
GET /search?q=... Substring search across title, description, attack class, behavioral fingerprint
GET /docs Interactive OpenAPI docs (auto-generated)

No authentication. No write endpoints. Rate limited to 60 requests/minute per IP as a basic abuse backstop, generous for any legitimate use.

Running locally

pip install -r requirements.txt
uvicorn main:app --reload

Data source

Fetches dist/ave-records-latest.json from aveproject/ave on startup and every 15 minutes thereafter. This repo holds no copy of the record data itself; aveproject/ave is the single source of truth.

Deployment

Deployed on Render. See ARCHITECTURE.md's ADR section for why Render was chosen over Cloud Run and Vercel, and what would have to change to revisit that.

Contributing

See CONTRIBUTING.md.

Security

See SECURITY.md.

License

Apache 2.0, see LICENSE.

About

Read-only reference API for AVE, the behavioral classification standard for agentic AI components. Not the standard itself, see aveproject/ave.

Topics

Resources

License

Code of conduct

Contributing

Security policy

Stars

1 star

Watchers

0 watching

Forks

Releases

No releases published

Packages

 
 
 

Contributors