Low Severity
Published: Oct 25, 2026

[GitHub] AgentDossier

Dos
Type
N/A
CVSS Score
3
Views
Anonymous
Author

Description

AgentDossier - A vendor-neutral registry of AI agents that shows what compliance and trust evidence exists for each Agent (HIPAA, SOC 2, FedRAMP, ISO42001, EU AI Act), Traces and CVEs. Agents are discovered through the open standards they publish (ARD, A2A, MCP) as well as public registries, and same engine can scan enterprise's own network.

# AgentDossier

**An evidence dossier for every AI agent.** Discover AI agents. Inspect the evidence. Decide what fits your risk requirements.

AgentDossier is a vendor-neutral registry of AI agents that leads with
evidence instead of marketing. For each agent it shows what compliance and
trust evidence exists — HIPAA, SOC 2, FedRAMP, GDPR, ISO/IEC 42001, the EU AI
Act and more — who issued it, what it covers, whether it is still current and
what you still have to verify yourself. Agents are discovered through the open
standards they publish (Agentic Resource Discovery, Agent2Agent, Model Context
Protocol) as well as public registries and marketplaces, and the same engine
can scan an enterprise's own network for the agents already running there.

[![AgentDossier homepage: find agents you can trust, check the evidence, see how they behaved under attack — search, a live dossier specimen with its trace, and the newest incidents](docs/screenshots/home-search.png)](https://sushegaad.github.io/AgentDossier/)

- **Public demo:** https://sushegaad.github.io/AgentDossier/ — static, no accounts, no cookies; the demo build sends only aggregate, cookieless page counts
- **Self-hosted edition:** the same MIT codebase as one container with an authorized private-network scanner, ARD REST API, policy qualification, decision workflow and integrations
- **Worked comparison:** [three ways to let a coding agent into a private repository](https://sushegaad.github.io/AgentDossier/compare/coding-agents-private-repo/) — the same five deployment questions answered for six agents from their own documentation
- **Docs:** [`docs/methodology.md`](docs/methodology.md) (how evidence is graded), [`docs/traces.md`](docs/traces.md) and [`docs/traces-spec.md`](docs/traces-spec.md) (traces and vendor publishing), [`docs/incidents.md`](docs/incidents.md) (the incident registry and its weekly upkeep), [`docs/cves.md`](docs/cves.md) (the CVE registry), [`docs/enterprise-runbook.md`](docs/enterprise-runbo

Community Rating
0

Login to rate this exploit

Quick Actions