claude-md-improver
Pass
Audited by Gen Agent Trust Hub on Aug 19, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection via untrusted repository content.
- Ingestion points: The skill reads
CLAUDE.md,.claude.md, and.claude.local.mdfiles from the repository and the user's home directory. - Boundary markers: There are no explicit delimiters provided to distinguish file content from system instructions, creating a risk that embedded malicious prompts could be executed.
- Capability inventory: The skill utilizes
Bashfor command execution andEditfor file modification, and the quality criteria suggest the agent verify documented commands by running them. - Sanitization: No validation or filtering is performed on external content before it is processed by the agent.
Audit Metadata