publish-linear

Pass

Audited by Gen Agent Trust Hub on Mar 30, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill's primary function is to automate the transfer of project metadata and documentation to Linear, which is a legitimate and well-known project management service. All file operations are restricted to the local project structure (./plans and ./issues).
  • [DATA_EXFILTRATION]: While the skill transmits project data to an external API (Linear), this is the stated and intended purpose. The transmission is handled through platform-provided tools rather than arbitrary network calls. No access to sensitive system files, such as SSH keys or cloud credentials, was detected.
  • [PROMPT_INJECTION]: The instructions for the agent are procedural and focus on extraction and mapping (e.g., 'Extract title from heading', 'Parse status from table'). This design limits the risk of the agent misinterpreting content within the synchronized files as instructions to override its core logic or bypass safety filters.
  • [COMMAND_EXECUTION]: The skill does not contain any shell commands, system calls, or dynamic code execution patterns. It relies entirely on structured tool calls and localized file reading.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 30, 2026, 12:51 PM