notemdpro-link-analyzer
Pass
Audited by Gen Agent Trust Hub on Apr 16, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill presents an indirect prompt injection surface. It reads external markdown files and incorporates their contents into an LLM prompt. Because the resulting output is used to modify or create files, malicious instructions embedded in the source notes could influence the LLM to perform unauthorized file changes.
- Ingestion points: Content is ingested from user markdown files via the
read_filetool infileUtils.ts. - Boundary markers: The prompt template defined in
promptUtils.tsdoes not utilize delimiters (such as XML tags or unique tokens) to isolate the user-provided data from the system's instructions. - Capability inventory: The skill has the capability to modify existing files and create new concept notes using the
write_filetool. - Sanitization: There is no evidence of output validation or sanitization before the content returned by the LLM is written to the local disk.
Audit Metadata