hrt-artifact-lint
Pass
Audited by Gen Agent Trust Hub on Aug 30, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes markdown files from a user-provided directory and generates output intended for interpretation by other AI agents. Maliciously crafted content within these files (such as capability names) could be used to inject instructions into the calling agent's context.\n
- Ingestion points: The lint.cjs script reads content from proposal.md, design.md, tasks.md, and various spec.md files within the provided directory.\n
- Boundary markers: The script lacks explicit boundary markers or instructions to downstream agents to ignore embedded text when outputting findings.\n
- Capability inventory: The script is limited to reading files (fs.readFileSync) and printing to the console; it does not have network access or write permissions.\n
- Sanitization: Extracted data, such as capability names, are interpolated directly into error messages (e.g., 'capability "${name}" is not kebab-case') without sanitization or escaping.
Audit Metadata