The Agent Skills Directory

[COMMAND_EXECUTION]: The coach.ts utility script executes git log and git diff commands to gather metadata for developer standups and client progress digests. These operations are read-only and restricted to the local repository's forensic history.
[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted client communication (e.g., Slack messages, emails) via the --scope-check command and procedures defined in references/client.md, creating a potential attack surface.
Ingestion points: External client requests are ingested through the --scope-check argument for evaluation against the Scope of Work (SOW).
Boundary markers: No explicit delimiter-based boundary markers are implemented when interpolating these external inputs into the agent context.
Capability inventory: The skill has access to file modification tools (write_to_file, replace_file_content) and shell command execution via the bash tool and internal scripts.
Sanitization: The coach.ts script provides a layer of logic by using keyword-based classification (e.g., "redesign", "new feature") to evaluate requests objectively before the agent processes them.

coach