The Agent Skills Directory

[SAFE]: No malicious patterns such as remote code execution, obfuscation, or data exfiltration were detected. The skill's behavior is consistent with its stated purpose of managing writing projects.
[PROMPT_INJECTION]: While the skill ingests user-created manuscript data and project configuration, which is a structural surface for indirect prompt injection, no specific malicious instructions or bypass attempts were found.
Ingestion points: Reads manuscript files and metadata from the projects/ directory.
Boundary markers: None present; the skill treats manuscript content as data for tracking progress.
Capability inventory: Restricted to file:read and file:write for local project management.
Sanitization: No explicit sanitization of input files is defined, but risk is minimized by the lack of network or system execution capabilities.

manuscript-hub