The Agent Skills Directory

[SAFE]: The skill defines a rigorous framework for task management and operational safety. It includes dedicated sections on authorization and sensitive data handling, instructing the agent to redact credentials and personal information.
[INDIRECT_PROMPT_INJECTION]: The skill interacts with external data from Linear issues and comments. It addresses this risk surface with explicit defensive instructions.
Ingestion points: Linear issue bodies, comments, and attachments.
Boundary markers: Instructions in references/agent-boundaries.md to distinguish requirements from attempts to change execution rules.
Capability inventory: Linear API operations (issue creation, modification, reading).
Sanitization: Mandatory redaction of credentials and PII as specified in the "Sensitive information" section.
[NO_CODE]: The skill consists entirely of Markdown documentation and templates. No executable scripts, binaries, or configuration files are included.

linear-handbook