tome
Pass
Audited by Gen Agent Trust Hub on Oct 18, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [SAFE]: No malicious patterns or security violations were found in the skill's instructions or reference documentation.
- [INDIRECT_PROMPT_INJECTION]: The skill possesses an attack surface for indirect prompt injection as it processes untrusted inputs from git history and external drafts. 1. Ingestion points: Git diffs, PR descriptions, and user-provided transcripts in SKILL.md and article-handoffs.md. 2. Boundary markers: Explicit instructions to label inferences with '[Inference: evidence]' and separate fact from interpretation in SKILL.md. 3. Capability inventory: Documentation generation, file creation, and execution of platform-specific CLI tools like 'zenn-cli' in SKILL.md and article-platform-optimization.md. 4. Sanitization: Instructions to mask sensitive internal details and unreleased features in SKILL.md and article-patterns.md.
- [COMMAND_EXECUTION]: The skill utilizes standard developer commands including 'git reflog' and 'npx zenn-cli' to fulfill its purpose of managing repository documentation and publishing articles. Evidence found in SKILL.md and article-platform-optimization.md.
Audit Metadata