Import unseen Teams meeting transcripts from SharePoint Stream or Clipchamp into the local Codex transcript workspace. Use when Codex needs to open the Stream video hub, filter to Meetings, inspect individual meeting detail pages, download transcript exports as `.vtt` and `.docx`, stage them under `/Users/tfinklea/codex/.tmp/transcript-imports`, ingest the canonical Teams transcript into `codex transcripts`, and record permission or export blockers for meetings that cannot be imported.
Use Playwright MCP for every browser action and the bundled helper script for all local state, download claiming, ingestion, and final reporting.
/Users/tfinklea/codex./Users/tfinklea/codex/.tmp/transcript-imports as the only staging area for raw downloads./Users/tfinklea/codex/transcripts.Set the helper path once:
export STREAM_IMPORTER="$HOME/.codex/skills/teams-stream-transcript-import/scripts/stream_meeting_import.py"
Only set a cutoff when the user explicitly wants a time-bounded pass. Do not assume that older meetings are already processed just because they are older or because a prior helper run marked them as cleared by cutoff.
python3 "$STREAM_IMPORTER" start-run
python3 "$STREAM_IMPORTER" apply-cutoff \
--run-id "<run-id>" \
--on-or-before YYYY-MM-DD