The Agent Skills Directory

[SAFE]: The skill acts as a synthesis tool that converts existing conversation context into a structured document (PRD) using a provided template.
[SAFE]: It employs security-best-practice defaults by storing output in a local .scratch directory and explicitly stating that hosted trackers from the Git remote will not be used without the user's explicit approval.
[SAFE]: Repository exploration and the use of a domain glossary are restricted to local context, which is consistent with the skill's primary purpose of synthesizing available information.
[SAFE]: The skill references local metadata (e.g., docs/agents/triage-labels.md) to resolve labels, avoiding hardcoded or unverifiable external configurations.

to-spec