qa-feature

Pass

Audited by Gen Agent Trust Hub on Aug 26, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTIONDATA_EXFILTRATION
Full Analysis
  • [COMMAND_EXECUTION]: The skill workflow involves executing scripts found in the workspace's package.json file. This allows for the execution of arbitrary shell commands if the repository configuration is malicious.
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection due to its processing of untrusted data from GitHub and Notion. 1. Ingestion points: Pull request titles, descriptions, comments, and diffs, as well as Notion ticket requirements. 2. Boundary markers: The instructions lack delimiters or directions to ignore instructions found within the external data. 3. Capability inventory: Access to GitHub and Notion via MCP, file system read access, and command execution. 4. Sanitization: No sanitization is specified for the ingested content.
  • [DATA_EXFILTRATION]: The skill reads from private GitHub and Notion sources, which could lead to exposure of sensitive project information if the agent's instructions are subverted.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 26, 2026, 04:26 PM
Security Audit — agent-trust-hub — qa-feature