prepare-pr

Pass

Audited by Gen Agent Trust Hub on Feb 23, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill uses local scripts (scripts/pr-prepare and scripts/committer) to manage the PR lifecycle. These scripts handle environment setup, execution of validation tests (gates), and pushing changes to the remote repository. These are standard automation components within this developer workflow.
  • [PROMPT_INJECTION]: The skill processes data from .local/review.json which contains findings and suggested fixes from external PR reviews. This creates a surface for indirect prompt injection.
  • Ingestion points: Data is ingested from .local/review.json using jq commands in SKILL.md.
  • Boundary markers: No explicit boundary markers or 'ignore embedded instructions' warnings are present when the agent is directed to resolve findings.
  • Capability inventory: The agent has the ability to execute repository-level scripts (scripts/pr-prepare, scripts/committer) and perform Git operations (commit, push).
  • Sanitization: No explicit sanitization or validation of the content within the JSON review file is performed before the agent acts on the 'fix' instructions.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 23, 2026, 12:56 PM