pre-changelog

Pass

Audited by Gen Agent Trust Hub on Aug 7, 2026

Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [PROMPT_INJECTION]: The skill processes content from pull request diffs fetched via the GitHub API, which presents a surface for indirect prompt injection where malicious text in a PR could attempt to influence agent behavior. \n
  • Ingestion points: Pull request diff data for apps/kimi-code/CHANGELOG.md is fetched in the workflow. \n
  • Boundary markers: No explicit delimiters or instructions are used to treat the fetched content as untrusted data. \n
  • Capability inventory: The skill uses gh for data retrieval; no capabilities for writing files, executing code, or privilege escalation were identified. \n
  • Sanitization: The skill applies specific transformation rules (stripping, merging, and translating) that limit how the input text is interpreted and used.\n- [COMMAND_EXECUTION]: The skill utilizes the GitHub CLI tool (gh) to perform read-only operations such as listing pull requests and querying repository API data within the MoonshotAI/kimi-code repository to retrieve changelog content.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 7, 2026, 07:30 PM
Security Audit — agent-trust-hub — pre-changelog