change-clarify

Pass

Audited by Gen Agent Trust Hub on Aug 20, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill is designed to manage local development metadata within the .adlc directory, with no external network access or credential usage.
  • [COMMAND_EXECUTION]: The skill executes local setup scripts (setup-change-clarify.sh and setup-change-clarify.ps1) to count and list pending records. These scripts are self-contained, perform only basic filesystem operations, and do not accept untrusted external input.
  • [DATA_EXFILTRATION]: The instructions include a security-focused requirement to paraphrase issue summaries instead of pasting them verbatim, which mitigates the risk of accidental sensitive information leakage.
  • [PROMPT_INJECTION]: The skill implements a provenance check, requiring every decision claim to cite a commit SHA or issue URL. This acts as a circuit breaker against context poisoning by ensuring decisions are grounded in verifiable history rather than hallucinated or injected content.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 20, 2026, 02:08 PM
Security Audit — agent-trust-hub — change-clarify