tech-doc-style-chinese
Pass
Audited by Gen Agent Trust Hub on Oct 10, 2026
Risk Level: SAFE
Full Analysis
- [EXTERNAL_DOWNLOADS]: The README.md provides instructions to install the skill and fetch updates from a public repository hosted on GitHub (github.com/Fenng/tech-doc-style-chinese), which is a well-known service.
- [COMMAND_EXECUTION]: The README.md instructs users to execute shell commands such as 'npx skills add', 'git clone', and 'python scripts/lint_copy_rules.py' for installation and stylistic validation.
- [INDIRECT_PROMPT_INJECTION]: The skill instructions in SKILL.md target technical document rewriting, creating an ingestion surface for untrusted text. 1. Ingestion points: Text inputs for rewriting in SKILL.md and NoCode-Skill.md. 2. Boundary markers: Partially present through instructions defining exclusions for code blocks and machine-readable tags. 3. Capability inventory: Limited to text formatting; no independent network or file-write capabilities across all scripts. 4. Sanitization: Absent.
- [DYNAMIC_EXECUTION]: The included Python script (scripts/lint_copy_rules.py) is executed to lint documentation. Analysis of the script confirms it only performs local read-only operations on Markdown files for regex matching and does not use dynamic evaluation (eval/exec) or process remote content.
Audit Metadata