Agent-Field/sec-af

AI-native security auditor on AgentField that proves exploitability with verdicts, traces, and actionable evidence.

36
/ 100
Emerging

This project utilizes an adversarial agent architecture, employing separate "hunter" agents to find vulnerabilities and "prover" agents to rigorously verify them through a 4-agent chain, including taint tracing and exploit hypothesis generation. It filters findings in stages, progressively reducing 106 raw findings to 30 confirmed vulnerabilities by applying semantic similarity for deduplication and deep data flow analysis for verification. The auditor integrates with and targets the AgentField framework, processing repositories via a single API call for comprehensive security audits.

No Package No Dependents
Maintenance 13 / 25
Adoption 6 / 25
Maturity 9 / 25
Community 8 / 25

How are scores calculated?

Stars

22

Forks

2

Language

Python

License

Apache-2.0

Last pushed

Mar 12, 2026

Commits (30d)

0

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/agents/Agent-Field/sec-af"

Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.