hmos-arkts-knowledge-retriever

Pass

Audited by Gen Agent Trust Hub on Aug 2, 2026

Risk Level: SAFE
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill includes instructions to download a custom TypeScript runtime from GitCode (gitcode.com/openharmony), which is a well-known service providing official source code for the OpenHarmony project. This process is documented as a standard setup step for the bundled linter tool.
  • [COMMAND_EXECUTION]: The retrieval workflow involves executing a local Python script (scripts/search_docs.py) and a Node.js-based linter CLI (linter-cli/bin/linter-cli.js). These commands are used for local documentation indexing and static code analysis, following best practices for development-oriented AI agent skills.
  • [PROMPT_INJECTION]: The skill's instructions in SKILL.md emphasize grounding responses in local repository sources and distinguish between implementation-derived evidence and official specification text. There are no attempts to bypass safety filters or override agent constraints.
  • [DATA_EXFILTRATION]: No evidence of data exfiltration was found. The skill does not access sensitive system files or environment variables and does not initiate network connections to untrusted domains.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 2, 2026, 05:44 PM
Security Audit — agent-trust-hub — hmos-arkts-knowledge-retriever