piiiico/a2a-trust-audit-action
Audit your A2A agent card across L1-L4 trust dimensions on every PR. GitHub Action wrapping @agentlair/a2a-trust-audit.
SUMMARY AI要約 by gpt-5-mini
A2A Trust Audit is a GitHub Action that audits A2A agent cards across four trust layers (L1 Identity, L2 Authentication, L3 Authorization, L4 Behavioral Trust) on every PR or push. Intended for maintainers and CI pipelines of A2A agents (e.g., AgentLair users), it runs 22 weighted checks (identity, DID, signed card, JWKS, x402/mTLS, capabilities, attestation, audit trail, etc.), computes per-layer scores, an overall numeric score and A–F grade, and emits workflow outputs (`grade`, `score`, `l1`–`l4`). Key features: in-place PR comment and job summary, configurable fail-on threshold, live README badge, self-contained vendored audit logic (no npm install), and read-only probing of well-known card endpoints. MIT licensed.
GitHub Topics
使用言語(バイト数比)
オーナー情報
日付
| GitHub作成日 | 2026-05-09 |
| 最終Push | 2026-05-09 |
| 当サイト初検出 | 2026-05-09 |
| 最終取得 | 2026-05-09 18:20 |
類似Repo (同じ言語のAI関連Repo)
一键检测浏览器环境,看 Claude Code 是否判定你为中国用户。纯前端,100% 本地运行,不上传任何数据。
Atulsharma8790/bug-hunter-agent CustomDigitalServices-Kevin/claude-config-studioWeb generator for .claude configurations (CLAUDE.md, settings.json, rules, hooks) with verified Claude Code marketplaces and third-party tools catalogs. 100% static, in-browser.
Bendak/hermes-agentosWeb UI / control plane for Hermes Agent — dashboard, kanban, sessions, and more.
gediminasurvila/video-editor-ai-electron BangDills/agent-relayA local-first CLI that compiles repository state into a verified handoff package for AI coding agents.
Jorgeotero1998/AihubUn “Operations OS” para empresas: agentes especializados, documentos con RAG, automatizaciones, calendario, reportes y realtime.
surrvezh/finance-trackerPersonal finance tracker — income, expenses, investments & AI insights