Knowledge

Source Library Ingestion QA

Turn raw captures into retrieval-ready knowledge instead of a pile of orphaned links.

Use when

A new article, video, X post, podcast, or image enters your knowledge base and should be findable later.

difficultyIntermediate
categoryKnowledge
sourcePaul-created

Cadence

After each source capture

Verification

Metadata complete, transcript/article state honest, useful takeaways present, and qmd retrieval verified or refreshed.

Structured loop spec

FieldValue
NameSource Library Ingestion QA
CategoryKnowledge
TriggerAfter each source capture
ObjectiveTurn raw captures into retrieval-ready knowledge instead of a pile of orphaned links.
Allowed inputsRelevant files, source notes, logs, tests, screenshots, metrics, or task state for this loop
Allowed actionsCheck canonical URL, title, author, platform, date, source type, and status.; Patch missing takeaways and best-practice sections from the transcript or article only.; Update the source index and refresh retrieval.; Stop when the note can be found by a distinctive query and has no avoidable TBD fields.
VerificationMetadata complete, transcript/article state honest, useful takeaways present, and qmd retrieval verified or refreshed.
Stop conditionStop when the verifier passes, the budget is exhausted, no progress is made, a blocker appears, or approval is required.
BudgetSet a time, turn, token, retry, file, or dollar cap before running the loop.
Approval boundaryHuman approval required before public, destructive, financial, legal, account, or production-impacting actions.
Safe outputDraft, report, checklist, table, or approval-gated recommendation
Works withClaude, ChatGPT, Gemini, any tool-using AI assistant

Steps

  1. Check canonical URL, title, author, platform, date, source type, and status.
  2. Patch missing takeaways and best-practice sections from the transcript or article only.
  3. Update the source index and refresh retrieval.
  4. Stop when the note can be found by a distinctive query and has no avoidable TBD fields.

Prompt

After a new source note is added, make it retrieval-ready. Verify source metadata, transcript or article state, key takeaways, extracted best practices, and search retrieval. Patch only what is grounded in the source. Output the note path, status, and retrieval proof.

Run in Claude Code

Paste this into Claude Code (or any tool-using agent) to run the loop bounded: one change per round, the same verification every round, durable state files, and explicit stop conditions.

Run the "Source Library Ingestion QA" loop from AI Loop Library (https://ailooplibrary.com/loops/source-library-ingestion-qa/) as a bounded loop.
Goal: Turn raw captures into retrieval-ready knowledge instead of a pile of orphaned links.
Rules: one change per round; run the same verification every round (Metadata complete, transcript/article state honest, useful takeaways present, and qmd retrieval verified or refreshed.); append each round to docs/loops/source-library-ingestion-qa/progress.md and update docs/loops/source-library-ingestion-qa/state.json; stop on verifier pass, 8 rounds, 3 consecutive failed verifications, no progress, a blocker, or anything needing human approval (money, production, outbound, deletion). Finish with a proof report: rounds used, changes made, verification output, remaining risk, and the next human decision.

Get the MCP server + agent pack

Tags

Obsidianretrievalresearchmetadata

Related loops

Browse all 68 loops