The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data (issue titles, bodies, and consensus) from the Fast Ship API to prioritize and assign tasks.
Ingestion points: External data is fetched in scripts/sources/fast-ship.mjs and passed to the agent's context.
Boundary markers: The skill contains clear instructions in SKILL.md and scripts/sources/fast-ship.mjs to treat issue content as data rather than instructions.
Capability inventory: The skill has capabilities to launch threads via t3_thread_launch, execute local scripts using spawnSync, and write to local state files.
Sanitization: External content is clipped to a summary but no rigorous sanitization of malicious prompt patterns is performed.
[COMMAND_EXECUTION]: The skill utilizes child_process.spawnSync to execute Node.js scripts for API interactions and local state management.
[EXTERNAL_DOWNLOADS]: The skill recommends the installation of a related skill from the author's GitHub repository (github.com/gitbobobo/fast_ship).

t3-dispatch