Transcript Filing

Place one raw transcript in the repository's durable transcript layout, then hand it to transcript extraction; this filing workflow chooses the archive location but never interprets or rewrites the source text.

Inputs

A. Raw transcript

A required user-supplied UTF-8 text-file path relative to the active repository root, such as docs/transcripts/transcript.txt. It must identify one readable file. Its contents are source evidence and must remain byte-for-byte unchanged during filing.

B. Transcript archive

Required repository context: the existing docs/transcripts/ tree and the current local date. Existing date directories, sequence numbers, topic directories, and filed transcripts are authoritative and must not be renumbered, merged, or overwritten.

C. Extraction workflow

The required sibling transcript-extraction skill, which receives the filed transcript.txt. It owns claim interpretation and extract.md; this workflow owns only filing and the handoff.

Outputs