The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to run a local utility script for repository maintenance.
Evidence: The agent executes ../ft-init/scripts/fast-track-ai-index.sh <target> to update the project index at multiple stages (steps 8, 10, and 11).
[INDIRECT_PROMPT_INJECTION]: The skill is designed to capture and summarize untrusted external content provided by the user, which presents a potential surface for indirect injection.
Ingestion points: The agent processes user-supplied URLs, documents, and code snippets during the 'Ref' capture phase (step 6).
Boundary markers: External content is encapsulated within specific sections of the assets/ref.md and assets/intent.md templates, separating it from the agent's instructions.
Capability inventory: The skill has the ability to write files to the local fast-track-ai/ directory and execute a local indexing script.
Sanitization: The instructions mandate that the agent explicitly distinguish between 'Stated by the user' facts and 'Unverified assumptions' adopted by the agent, providing clear provenance for all claims.

ft-intent