TLS certificate checks disabled
200 of 477 eligible files · Partial selectionOct 15, 2026 · UTC
Metal Mantra is a public registry of automated reports on AI agents. Agents act on your behalf, so every report makes their public source findings, scan scope and unknowns legible in one place.
Tell us the job. Explore agents whose public repository descriptions may fit it.
mhtsec · Agent Signal v0.2
MAJOR · SEC005
Every report belongs to a specific method and moment. Start with what was found, then inspect what was checked.
Explore all reportsTLS certificate checks disabled
200 of 477 eligible files · Partial selectionA CLI that uses a DeepAgents documentation agent to generate and maintain an OpenWiki for a codebase.
Domain ownership not verified
200 of 231 eligible files · Partial selectionA multi-agent offensive-security framework, built to turn the AI coding agent you already run into a zero-day hunter.
View sourceHardcoded credential
186 of 189 eligible files · Partial selectionMetal Mantra makes the boundaries visible. Three different questions deserve three different answers.
Automated rules flag source and domain patterns. Follow each finding to its location, impact and suggested fix.
DNS verification demonstrates control of a domain. Entity trust contributes 15% of the v0.2 score; it does not certify security.
A source scan does not establish runtime safety, agent capability or fitness for your use case. Missing evidence stays visible.
Your first scan is free and becomes public when it finishes. Understand the findings, verify domain ownership and share the evidence.
Findings on public repositories are shown in full, with the line they came from, so you can judge the risk yourself. Secret values are never printed.
What Metal Mantra is, what a report can and cannot tell you, and how to use it. The same answers are in the user guide.
Metal Mantra is a public registry of automated evidence reports on AI agents. For each agent it shows what deterministic static rules found in the public source, what the owner proved about their domain, and what could not be checked. It is built by Danhoangda, an independent AI lab in Hanoi.
Most directories list what an owner says about an agent. Metal Mantra reads the public source with fixed rules and publishes the evidence: each deduction points to a rule and a file, and every report states how many files were eligible, how many were read and what stays unknown.
A score and grade across five weighted pillars (security and privacy 25%, guardrails and loop control 25%, schema and tooling 20%, token and cost efficiency 15%, entity trust 15%), the findings with the rule and file behind each one, the scan scope, domain evidence marked as verified or not, and a plain list of what the report does not establish.
No. A score is an automated signal about a bounded snapshot of public source. It does not establish runtime safety, real-world capability or fit for a particular job, and every report says so.
No. Scores come from deterministic rules, so the same repository snapshot gives the same result. AI is used in one optional place only: matching a buyer's plain-language job description to agents whose public descriptions may fit. It never changes a score.
Yes. The first scan of a supported public GitHub repository is free and becomes a public report. A re-audit pass costs $29 for three re-scans of one repository. There is no subscription, and payment never changes a score or a ranking.
Yes, without sharing the source. Run the Metal Mantra scanner inside your own GitHub Actions job. It sends rule identifiers and counts, authenticated by GitHub's OIDC token, and the report is labelled CI-attested, which is visibly weaker than a scan Metal Mantra runs itself.
Yes. A read-only MCP server at https://metalmantra.io/mcp needs no sign-in. Its tools include search_agents, get_agent_report, verify_trust and get_standard. Call verify_trust with an agent's owner/repo before delegating work to it: "unknown" means not checked, never a pass.
The method is public: every rule, severity and weight is listed on the Standard page. The scanner's code is not open source today.