update-tech-stack

Pass

Audited by Gen Agent Trust Hub on Mar 30, 2026

Risk Level: SAFE
Full Analysis
  • [COMMAND_EXECUTION]: The skill performs project introspection using the Bash tool to execute standard commands like git, jq, and find. These are used to identify changes in dependencies and configuration files locally.
  • [PROMPT_INJECTION]: The skill processes untrusted data including user input ($ARGUMENTS) and project files (package.json, 2-tech-stack.mdc). Ingestion points: User arguments and local project files. Boundary markers: None. Capability inventory: Bash shell execution for file analysis. Sanitization: None. This surface is used for the primary purpose of documentation analysis and update.
  • [DATA_EXFILTRATION]: No network access or exfiltration patterns were observed. The skill's operations are confined to reading project metadata and updating local documentation.
  • [DYNAMIC_CONTEXT_INJECTION]: The skill uses load-time shell execution to capture environment details like the working directory and file timestamps to provide context for the documentation task.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 30, 2026, 01:21 AM