The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from external trackers, which serves as a potential vector for indirect prompt injection if malicious instructions are embedded in ticket titles, bodies, or comments.
Ingestion points: The agent retrieves data from GitHub Issues, Linear tickets, and local markdown files, including issue bodies and comments.
Boundary markers: The instructions do not define specific delimiters or isolation techniques to separate untrusted issue content from the agent's primary instruction set.
Capability inventory: The skill utilizes the gh (GitHub) CLI, Linear MCP tools, and local filesystem writes to manage ticket states.
Sanitization: The skill follows security best practices by instructing the agent to use --body-file and JQ parsing in github.md, which significantly reduces the risk of shell-based command injection from untrusted issue content.
[COMMAND_EXECUTION]: The skill executes local commands using the gh CLI and potentially git to interact with the repository's ticket tracking configuration and GitHub's API.
[DATA_EXFILTRATION]: The skill transmits issue data and metadata to GitHub and Linear. These operations target well-known service providers and are essential to the skill's documented functionality for synchronizing ticket states.

tracker