debug-access-token
Pass
Audited by Gen Agent Trust Hub on Oct 30, 2026
Risk Level: SAFE
Full Analysis
- [DATA_EXFILTRATION]: The skill implements robust security measures to prevent access tokens and app secrets from entering the agent's context. It provides a local Python script,
scripts/debug_token_probe.py, which is designed to be run in the developer's own environment to process credentials and only return non-sensitive, redacted metadata. - [EXTERNAL_DOWNLOADS]: The skill references an official Model Context Protocol (MCP) server at
https://mcp.facebook.com/devtoolsfor documentation discovery. This is a well-known service provided by Meta and is used according to its intended purpose. - [COMMAND_EXECUTION]: The skill provides shell instructions for the developer to run locally. These instructions include security-conscious practices, such as using
read -sto prevent tokens from appearing in shell history and using environment variables to keep secrets out of process listings. - [SAFE]: No malicious patterns, obfuscation, or unauthorized data access behaviors were detected. The skill's architecture correctly shifts sensitive operations to the user's local environment, maintaining a high security posture.
Audit Metadata