Turn Sapat transcripts into a knowledge base

[Mykidzou]

/claim #13

Summary

• Add a distinct long-form Daytona guide for turning Sapat transcript output into a structured transcript knowledge base.
• Cover reproducible Daytona setup, Sapat provider configuration, single-file and directory transcription, correction pass tradeoffs, and Markdown note review workflow.
• Add an original SVG workflow diagram, a Daytona-style definition for transcript knowledge base, and a first-time contributor author profile.

Validation

• npx --yes markdownlint-cli guides/20260511_turn_sapat_transcripts_into_knowledge_base_daytona.md definitions/20260511_definition_transcript_knowledge_base.md authors/mykidzou.md --ignore node_modules
• git diff --check
• python -m pip install -e . in the current Sapat repository
• python -m compileall src in the current Sapat repository
• sapat --help
• Verified referenced public links that allow scripted fetch: Sapat, Daytona docs, Daytona CLI docs, Groq speech-to-text docs, and Azure OpenAI audio docs. The official OpenAI docs link is valid in browser/search but returns 403 to scripted fetches from this environment.

Notes

This submission is intentionally scoped away from the recent provider-comparison, batch QA, privacy/redaction, and generic setup submissions. It focuses on post-transcription structuring: chapters, summaries, decisions, command references, open questions, and action items.

Transparency note: this guide was drafted with AI assistance, then checked against the current Sapat README/source and validated with the commands above.