The Agent Skills Directory

[SAFE]: The skill is entirely instructional and uses a provided markdown template. It contains no executable scripts or external dependencies.
[SAFE]: Access is restricted to reading transcripts and repository context (ADRs, existing docs) and writing to a local report file. No network capabilities or sensitive file access patterns are present.
[SAFE]: The process instructions emphasize traceability and evidence linking (verbatim quotes), which serves as a functional control against the agent misinterpreting or misrepresenting the source text.

transcript-extraction