sharemd
Pass
Audited by Gen Agent Trust Hub on Aug 1, 2026
Risk Level: SAFEDATA_EXFILTRATIONPROMPT_INJECTIONCREDENTIALS_UNSAFE
Full Analysis
- [DATA_EXFILTRATION]: The skill is designed to transmit markdown content or file data to the external service 'sharemd.dev'. This is the primary intended function provided by the vendor 'veronalabs'.
- [CREDENTIALS_UNSAFE]: The instructions demonstrate a safe practice by explicitly warning users not to paste API keys into the chat and directing them to use a secure Authorization header in the MCP connection configuration.
- [PROMPT_INJECTION]: The skill possesses an indirect prompt injection surface (Category 8) due to its file-sharing functionality.
- Ingestion points: The skill parses file paths provided in the conversation context to determine what data to share.
- Boundary markers: None. There are no instructions for the agent to verify that the file path is safe or to ignore instructions embedded within the files it reads.
- Capability inventory: The skill uses the 'mcp__sharemd__share-markdown' tool which reads data and performs a network POST request to an external domain.
- Sanitization: The skill does not perform any validation or filtering on the content retrieved from the files before transmitting it to the public link.
Audit Metadata