woostack-commit

Pass

Audited by Gen Agent Trust Hub on Aug 29, 2026

Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTIONDATA_EXFILTRATION
Full Analysis
  • [COMMAND_EXECUTION]: Executes Git, Graphite (gt), and GitHub (gh) CLI tools to perform repository state mutations and pull request updates.
  • [COMMAND_EXECUTION]: Invokes a local helper script 'scripts/change-receipt.sh' to audit repository state and allows execution of a configurable 'commit.command' from the local repository settings for pre-staging verification.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes potentially untrusted data from external artifact providers like Linear and Plane. Ingestion points: issue identifiers and metadata; Boundary markers: managed markers in PR bodies; Capability inventory: Git/Graphite/GitHub shell commands; Sanitization: Explicit instructions mandate treating external content as untrusted to prevent injection.
  • [DATA_EXFILTRATION]: Transmits code diffs and task metadata to GitHub and third-party issue tracking providers as a functional requirement for PR association and delivery notes.
  • [SAFE]: Implements a strict safety policy in the staging phase to exclude secrets, environment files (.env), and other sensitive local configuration.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 29, 2026, 04:42 AM
Security Audit — agent-trust-hub — woostack-commit