The Agent Skills Directory

[DATA_EXFILTRATION]: The skill facilitates the transmission of local file content and remote URL data to the Mowen backend. This provides a mechanism for exfiltrating sensitive local files (e.g., SSH keys, .env files, or cloud credentials) if the agent is directed to unauthorized paths. Evidence: The mocli misc upload --file command in SKILL.md transmits local filesystem content to a remote server.
[PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface. External data processed by the agent could contain hidden instructions to trigger the file upload functionality on sensitive local file paths. Ingestion points: Local file paths and remote URLs supplied via agent context or user input. Boundary markers: The skill instructions advise the agent to confirm before uploading sensitive or private information, providing a procedural guardrail but lacking strict technical boundary enforcement. Capability inventory: The mocli misc upload tool has the capacity to read local files and perform network operations. Sanitization: There is no evidence of path or URL validation to prevent access to sensitive system directories.
[COMMAND_EXECUTION]: The skill requires the execution of a vendor-specific CLI tool to perform its primary tasks. Evidence: Use of the mocli binary is defined in the skill's metadata and instructional body.

mo-misc