OVERVIEW 82% of enterprises report 'unknown agents' and 65% have had an agent-related incident. Moltline Agent Governance is the audit toolkit for the fleet you actually run: paste an MCP config, a tool manifest, a skill file, or an inventory and get a risk-ranked report with a concrete remediation per finding.

KEY CAPABILITIES • audit_mcp_config — leaked secrets, ungated credentials, unpinned launchers, auto-approve flags → posture score + fixes • scope_check — blast-radius ranking of a tool manifest (command exec > money/delete > write > read) • audit_skill_file — injection, guardrail-bypass, concealment, and exfiltration smells in third-party skills • injection_scan — screen untrusted content before it reaches an agent's context • inventory_report — tiered governance inventory; flags unowned agents • Pro: governance_policy (audit-ready, fleet-tailored) + get_auditor_persona

USE CASES • Vet a third-party MCP server or skill before installing it • Build the agent inventory your security review keeps asking for • Screen scraped web content for prompt injection before ingestion

WHO THIS IS FOR Platform and security engineers governing agent fleets, teams adopting MCP at scale, and anyone who installs third-party skills and wants to know what they actually do.

WHY MOLTLINE • Every tool tested against known answers before release • One MOLT- license unlocks premium across the whole Moltline line • Support: first response <24h, confirmed defects fixed <30 days

Independently audited — MCPize Verified A on the direct endpoint https://mcp.moltlinestudio.com/govern · Live check result