The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes a bundled shell script scripts/tickets.sh to perform issue tracking operations. This script invokes the gh CLI for GitHub interactions and executes a static, embedded Python script to interface with Linear and Jira APIs.
[INDIRECT_PROMPT_INJECTION]: The skill ingests and processes untrusted data from issue bodies, comments, and pull request diffs which are controlled by external users.
Ingestion points: Issue data is gathered via gh issue view and the issue (Linear) or view (Jira) API calls within the scripts/tickets.sh utility.
Boundary markers: While the skill applies a disclaimer to its own generated comments, it lacks explicit boundary markers or "ignore" instructions when interpolating retrieved issue content into its reasoning context.
Capability inventory: The skill has the ability to modify issue metadata (labels), post comments as the authenticated user, and close tickets.
Sanitization: No explicit sanitization or filtering of the external issue content is performed before the agent analyzes it to recommend triage states, creating a potential vector for malicious issue descriptions to influence agent actions.
[CREDENTIALS_SAFE]: Sensitive information such as LINEAR_API_KEY and JIRA_API_TOKEN is handled through environment variables. The skill explicitly instructs the agent to never hardcode or write credentials to files, adhering to security best practices for secret management.
[EXTERNAL_DOWNLOADS]: The skill makes network requests to well-known and trusted services including api.linear.app and user-defined Jira Cloud instances (JIRA_BASE_URL). These connections are required for the skill's core functionality and are established using TLS via enforced https protocols.

sift