The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests data from codebase comments, which constitutes a surface for indirect instructions to influence the agent.
Ingestion points: The skill uses grep to extract text from every file in the repository that contains the ponytail: prefix.
Boundary markers: There are no boundary markers or instructions to treat the harvested comment text as data rather than instructions.
Capability inventory: The skill executes shell commands (grep, git blame) and has the capability to write a ledger to the filesystem (e.g., PONYTAIL-DEBT.md).
Sanitization: There is no evidence of sanitization or filtering applied to the harvested comment content before it is processed or written to output.
[COMMAND_EXECUTION]: The skill invokes shell utilities to perform its core functionality.
Evidence: The instructions explicitly direct the agent to run grep -rnE '(#|//) ?ponytail:' . and potentially git blame on specific lines to gather data from the local environment.

ponytail-debt