The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process information from various repository files such as ADRs, glossaries, and existing documentation to inform its specification-writing process. This creates a surface where malicious instructions embedded in those documents could theoretically influence the agent's behavior during the interview.
Ingestion points: Repository files, including docs/, ADRs, and glossary files, are read to gather context (SKILL.md).
Boundary markers: There are no explicit instructions for the agent to treat repository content as untrusted or to ignore instructions contained within those documents.
Capability inventory: The skill has the ability to read files using rg and write markdown files to the repository (SKILL.md).
Sanitization: No sanitization is performed on the ingested text before it is used to prompt the user or update the draft spec.
[COMMAND_EXECUTION]: The skill uses standard command-line tools to interact with the local repository environment.
Evidence: The instructions explicitly mention using rg --files to search for spec templates within the project directory.

grill-to-spec