Find exposed secrets, verify supported credentials with explicit consent, and act on them — in VS Code, on the command line, and over MCP. A GPY Analytics product.
LPM flags this version as an AI-agent control-surface risk. The MCP server supplies agent-facing instructions that demand its findings be treated as final. This can hijack an AI consumer's security judgment when the MCP server is connected.
Source executes local commands and sends command output to an external endpoint.
out/mcp.jsView on unpkg · L57A single source file combines environment access, network access, and code or shell execution; review context before blocking.
out/mcp.jsView on unpkg · L57Source reaches cloud instance metadata or link-local credential endpoints.
out/mcp.jsView on unpkg · L1Source contains an obfuscated payload loader that reconstructs and executes hidden code.
out/mcp.jsView on unpkg · L1A package entrypoint or install-time lifecycle script reaches a source file with blocking dangerous behavior.
out/cli.jsView on unpkg · L1A manifest entrypoint or package-local install chain reaches command-output exfiltration behavior.
out/cli.jsView on unpkg · L1A bounded semantic-analysis stage reached its safety limit; remaining detectors completed, but this package requires AI review.
package.jsonView on unpkgThis report applies to secretloop@0.5.0.
See version security history for other recorded verdicts.
Evidence last updated: .
Source executes local commands and sends command output to an external endpoint.
out/mcp.jsView on unpkg · L57A single source file combines environment access, network access, and code or shell execution; review context before blocking.
out/mcp.jsView on unpkg · L57Source reaches cloud instance metadata or link-local credential endpoints.
out/mcp.jsView on unpkg · L1Source contains an obfuscated payload loader that reconstructs and executes hidden code.
out/mcp.jsView on unpkg · L1A package entrypoint or install-time lifecycle script reaches a source file with blocking dangerous behavior.
out/cli.jsView on unpkg · L1A manifest entrypoint or package-local install chain reaches command-output exfiltration behavior.
out/cli.jsView on unpkg · L1A bounded semantic-analysis stage reached its safety limit; remaining detectors completed, but this package requires AI review.
package.jsonView on unpkg