post-implementation-reviewer

Pass

Audited by Gen Agent Trust Hub on May 15, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill instructs the agent to execute shell commands for auditing purposes. This includes using the GitHub CLI (gh) to list issues and labels, and running test suites using bun test or pytest. These are standard operations for a technical reviewer role and are performed within the scope of the project being audited.
  • [PROMPT_INJECTION]: The skill possesses an indirect prompt injection surface as it processes external, potentially untrusted data.
  • Ingestion points: Data is ingested from GitHub issue titles/descriptions (gh issue list) and external documentation files (e.g., docs/design/*.md).
  • Boundary markers: The instructions do not define specific delimiters to isolate external data from the agent's internal logic.
  • Capability inventory: The skill has capabilities to write to the file system (updating READMEs and guides), create GitHub issues, and execute code via test runners.
  • Sanitization: There is no explicit sanitization or validation logic for the content read from issues or design documents before it is used to modify other files in the repository.
  • [SAFE]: No obfuscation, data exfiltration to unknown domains, or persistence mechanisms were found. The skill's behavior aligns with its stated purpose of performing technical reviews.
Audit Metadata
Risk Level
SAFE
Analyzed
May 15, 2026, 05:43 AM
Security Audit — agent-trust-hub — post-implementation-reviewer