The Agent Skills Directory

[SAFE]: The skill is a standard workflow automation tool for the Linear platform. It contains no malicious patterns, such as credential harvesting, obfuscation, or unauthorized network operations.
[PROMPT_INJECTION]: The skill processes documents from the local thoughts/ directory to generate Linear ticket content (Ingestion point). It leverages MCP tools like mcp__linear__create_issue to interact with external services (Capability inventory). While explicit boundary markers and sanitization for ingested content are not defined in the instructions, the risk of indirect prompt injection is mitigated by a mandatory drafting and interactive refinement phase, requiring explicit user approval of the content before any API calls are executed.
[COMMAND_EXECUTION]: The skill mentions using tools like Grep to search local directories for relevant context. This is a legitimate functional requirement for identifying documentation relevant to a user's request and does not represent an arbitrary code execution risk.

linear