The Agent Skills Directory

[COMMAND_EXECUTION]: The skill invokes a shell script located in a sibling directory outside its own package. * Evidence: Step 9 executes ../ft-init/scripts/fast-track-ai-index.sh <target>. * Context: Path resolution rules indicate this script is external to the skill's directory, residing in a sibling ft-init component.
[DYNAMIC_EXECUTION]: The skill performs runtime execution of a secondary shell script to validate or index the repository. * Evidence: The skill triggers the execution of fast-track-ai-index.sh as a final lifecycle step.
[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from active conversations and external index files to generate structured handoff documents. * Ingestion points: Step 1 reads the current conversation history and the fast-track-ai/INDEX.md file. * Boundary markers: Step 2 and the assets/handoff.md template use the header ## Active user instructions, verbatim to delimit user-provided content. * Capability inventory: The skill can write files to the repository (Step 6) and execute shell scripts (Step 9). * Sanitization: Step 2 contains instructions to omit secrets or personal data, though this relies on the model's compliance rather than programmatic filtering.

ft-handoff