Lading-AI: audit your AI agents before you let them run

Lading-AI: audit your AI agents before you let them run is an AI tool for code workflows, especially connect mcp tools and build ai agents. A skill has executable instructions. An MCP config can reach a remote endpoint or spawn a local command. A chat takeout is often the most sensitive archive on disk. None of it was enumerable by reading config files. Lading ( Ai, not o11y { Datadog } ) is a static, fully offline scanner (no model, no network, no telemetry and network imports are banned by an. Verified public information highlights a skill has executable instructions, an MCP config can reach a remote endpoint or spawn a local command, a chat takeout is often the most sensitive archive on disk, none of it was enumerable by reading config files. In AI Finderz, the strongest matching use cases are connect mcp tools, build ai agents, improve developer productivity, search knowledge. The project is publicly presented as open source under MIT. The public project is open source under MIT. Check the official page for hosted services, support or usage-specific costs. AI Finderz checked the public source on 2026-09-24 and uses the official page as the factual basis for this profile.