The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: Fetches issue, project, and pull request data from GitHub and GitLab using official CLI tools. These operations target well-known services and are consistent with the skill's primary purpose.
[COMMAND_EXECUTION]: Executes shell commands via the Bash tool to interact with version control systems and local files.
Uses gh, glab, and git to retrieve remote repository data.
Uses grep to parse local files such as PLAN.md, tasks.md, and TODO.md for task items.
Performs write operations by posting comments to GitHub issues using the gh issue comment command.
Interpolates variables like $PROJECT_NUMBER, $OWNER, and $TASK_ID into shell commands; while these are typically sourced from internal policy or state, they represent potential injection points if the source data is manipulated.
[PROMPT_INJECTION]: The skill processes untrusted content from external issue trackers, creating a surface for indirect prompt injection.
Ingestion points: Issue titles and bodies fetched from GitHub and GitLab APIs, and content from local markdown files (SKILL.md).
Boundary markers: Absent; the skill does not use specific delimiters or instructions to prevent the agent from obeying instructions embedded within the fetched issue data.
Capability inventory: The skill has access to shell command execution (Bash), file system access (Read), and user interaction (AskUserQuestion).
Sanitization: The skill performs basic truncation on labels for UI compliance but lacks robust sanitization or validation of the fetched issue content before it is processed for scoring and display.

discover-tasks