[ { "id": "SOC01", "title": "Content Repurposing FACTORY", "creator": "Lucas Walter | AI Automation", "platform": "YouTube", "original_url": "https://www.youtube.com/watch?v=u9gwOtjiYnI", "companion_repo_or_walkthrough": [ "https://github.com/lucaswalter/n8n-ai-automations/blob/main/content_repurposing_factory.json", "https://github.com/lucaswalter/n8n-ai-automations" ], "observed_or_claimed_mechanics": "Static JSON reveals a two-branch draft generator, not demonstrated viral performance. Both prompts explicitly request invented metrics when source evidence lacks them.", "inputs": [ "YouTube URL", "English SRT subtitles", "Separate platform exemplar sets" ], "outputs": [ "Three Twitter options", "Three LinkedIn options", "Slack threads containing drafts and source link" ], "steps": [ "Form -> Apify synchronous subtitle fetch (300-second timeout).", "Extract first subtitle, title and URL.", "Branch into platform prompts with exemplar structures; Claude Sonnet 4 generates options.", "Schema parsers require exactly three objects per branch, with text and explanation.", "Split arrays; post options beneath platform-specific Slack parent messages." ], "quality_checks_known": [ "Static schemas constrain each branch to three outputs; they do not validate factual accuracy.", "No factual-check or approval node found; both prompts request fabricated quantitative impact." ], "missing_details": [ "Both prompts instruct fabrication of quantitative impact.", "Schema repair validates format, not truth.", "No social publishing, factual checker, or approval node found.", "First-subtitle access assumes subtitles exist.", "Sticky notes say two options; executable schemas specify three.", "Public JSON corrects prior community-only template limitation.", "Video fetch failed; 25+ assets claim unsupported by this graph." ], "candidate_patterns": [ "Cautionary source-to-audience adaptation mechanism; do not copy unchanged." ], "cross_function_transfer_hypothesis": "Hypothesis: adapt an approved product announcement into customer, sales, and internal versions while preserving factual claims.", "transfer_constraints": [ "Needs source-fidelity review", "Platform constraints", "Publication approval" ], "evidence_status": "creator-reported-or-static-artifact", "sponsorship_if_disclosed": "Promotes implementation services, community and newsletter; sponsorship not established.", "source_family": "Lucas Walter", "timestamps": [ "00:46 demo", "04:54 transcript extraction", "14:10 Twitter", "19:34 LinkedIn" ], "accessed_on": "2026-10-18", "independently_run": false, "video_observed": false, "timestamp_confidence": "Chapter locator from stated textual source; not checked against video playback", "original_video_or_post_publication_date_independently_verified": false, "date_provenance": { "type": "youtube_description_index", "locator": "Indexed YouTube metadata, not direct video inspection" }, "source_dates": { "published": "2025-07-09", "updated": null, "other_date": null, "other_date_type": null, "evidence_url": "https://www.youtube.com/watch?v=u9gwOtjiYnI", "evidence_note": "Initial public search index returned YouTube title, creator, description and Published: 2025-07-09. Direct video fetch failed. Exact-video-ID/date search on final check returned no results; initial indexed retrieval is retained, not upgraded.", "checked_at": "2026-10-18T03:00:07Z", "status": "provisional_indexed_original_metadata" }, "source_date_entries": [ { "published": "2025-07-09", "updated": null, "other_date": null, "other_date_type": null, "evidence_url": "https://www.youtube.com/watch?v=u9gwOtjiYnI", "evidence_note": "Initial public search index returned YouTube title, creator, description and Published: 2025-07-09. Direct video fetch failed. Exact-video-ID/date search on final check returned no results; initial indexed retrieval is retained, not upgraded.", "checked_at": "2026-10-18T03:00:07Z", "status": "provisional_indexed_original_metadata" }, { "published": null, "updated": null, "other_date": null, "other_date_type": null, "evidence_url": "https://github.com/lucaswalter/n8n-ai-automations/blob/main/content_repurposing_factory.json", "evidence_note": "Static public template retrieved. No creation/update timestamp established; current template may differ from original video.", "checked_at": "2026-10-18T03:17:43.174030+00:00", "status": "unknown_date" }, { "published": null, "updated": null, "other_date": null, "other_date_type": null, "evidence_url": "https://github.com/lucaswalter/n8n-ai-automations", "evidence_note": "README links exact original YouTube video to template; no README publication/update date established.", "checked_at": "2026-10-18T03:17:43.174030+00:00", "status": "unknown_date" }, { "published": null, "updated": null, "other_date": "2025-07-09T14:35:30Z", "other_date_type": "latest_file_commit_committer_timestamp", "evidence_url": "https://api.github.com/repos/lucaswalter/n8n-ai-automations/commits?path=content_repurposing_factory.json&per_page=1", "evidence_note": "GitHub commits API returns this latest file commit; author timestamp 2025-07-09T14:35:17Z. Commit message says adds workflow. Neither is a verified YouTube upload date.", "checked_at": "2026-10-18T03:18:20.211974+00:00", "status": "verified_repository_commit_timestamp" } ], "original_evidence_status": "Primary YouTube description indexed by search; direct video fetch failed; video not watched.", "evidence_interpretation": "Source documentation or reported practice only; no independent execution, effectiveness, safety, or transfer validation.", "reference_eligibility": { "eligible": true, "source_category": "firsthand_creator_build_or_practice", "reason": "Public creator-owned static template supports mechanism and a concrete failure finding: prompts instruct invented metrics. Eligible only for cautionary anatomy, never as a successful/ready-to-copy workflow.", "allowed_use": "cautionary_mechanism_only", "eligible_for_success_highlight": false, "effectiveness_proven": false, "independent_corroboration_status": "not_required_for_process_reference", "corroboration_links": [], "evidence_constraints": [ "Eligibility, where granted, permits a process reference only; it is not proof of usefulness, causal effectiveness, safety, or transferability.", "No workflow software executed, no video playback independently observed, and no outcomes independently measured by this audit.", "The exact public template is available and statically inspected; both platform prompts instruct fabrication of metrics. Availability does not make the template safe to reuse.", "No transcript inspected", "No performance evidence", "Both prompts instruct fabrication of quantitative impact.", "Schema repair validates format, not truth.", "No social publishing, factual checker, or approval node found.", "First-subtitle access assumes subtitles exist.", "Sticky notes say two options; executable schemas specify three.", "Video fetch failed; 25+ assets claim unsupported by this graph." ] }, "eligibility_register_id": "SOC01", "enrichment_applied": true, "enrichment_scope": "Additional inspected text/static artifacts; original historical record preserved in audit/input-snapshots/repo-handoff/social-builds.json.", "category": "content_adaptation", "mechanism": "Static JSON reveals a two-branch draft generator, not demonstrated viral performance. Both prompts explicitly request invented metrics when source evidence lacks them.", "evidence": [ { "url": "https://github.com/lucaswalter/n8n-ai-automations/blob/main/content_repurposing_factory.json", "inspection": "All nodes, parameters and connections read as static JSON; no workflow execution.", "locators": [ "form_trigger", "scrape_youtube_video", "set_youtube_details", "build_twitter_prompt: STEP 1, item 3", "build_linked_in_prompt: STEP 1, item 3", "tweet_parser", "linked_in_parser", "send_tweet_msg", "send_linked_in_post_msg" ] }, { "url": "https://github.com/lucaswalter/n8n-ai-automations", "inspection": "README explicitly links this filename to original video." } ], "limitations_and_failure_cases": [ "Both prompts instruct fabrication of quantitative impact.", "Schema repair validates format, not truth.", "No social publishing, factual checker, or approval node found.", "First-subtitle access assumes subtitles exist.", "Sticky notes say two options; executable schemas specify three.", "Public JSON corrects prior community-only template limitation.", "Video fetch failed; 25+ assets claim unsupported by this graph." ], "reusable_pattern_interpretation": "Interpretation: reusable source-to-audience adaptation requires an immutable fact sheet, separate style examples, and a claim-evidence review. Keep this as a cautionary example, not an unchanged starter.", "workflow_granularity": "Whole job: turn one source into reviewable channel drafts. Branches are reusable subworkflows; JSON parsing is a supporting step, not a separate user workflow.", "recommendation_status": "cautionary_do_not_copy_unchanged", "verified_template_version": { "commit_sha": "6729938f1ca6978fa9ddc20549a9805c24a2c58b", "commit_url": "https://github.com/lucaswalter/n8n-ai-automations/commit/6729938f1ca6978fa9ddc20549a9805c24a2c58b", "pinned_file_url": "https://github.com/lucaswalter/n8n-ai-automations/blob/6729938f1ca6978fa9ddc20549a9805c24a2c58b/content_repurposing_factory.json", "sha256": "e1b1e9a5944f22a1e849438a9376d2e051b2251b2d9f953ca9ba7600c7480eb8", "verification": "Pinned raw file bytes exactly match inspected main-branch file.", "license_and_redistribution": "Full upstream JSON retained as local research evidence only; do not include in deliverable/repository without separate license review." }, "steps_basis": "Static template inspection; not executed." }, { "id": "SOC02", "title": "Build the Context That Makes AI Data Agents Reliable", "creator": "Shane Butler / AI Analyst Lab", "platform": "YouTube + creator transcript", "original_url": "https://www.youtube.com/watch?v=LsaPQ5eQUyA", "companion_repo_or_walkthrough": [ "https://aianalystlab.ai/workshops/build-the-context-that-makes-ai-data-agents-reliable/", "https://maven.com/p/34e8f6" ], "observed_or_claimed_mechanics": "Creator reports a fresh-session retention comparison and shared YAML definitions; this session explicitly does not run the demo.", "inputs": [ "Metric question", "Warehouse", "Grain, numerator, denominator, time window and exclusions" ], "outputs": [ "Definition-aligned answer", "Clarification when meaning would fork" ], "steps": [ "Reported comparison: ask the same ambiguous metric question in fresh sessions", "Provide the metric definition", "Repeat and compare answers" ], "quality_checks_known": [ "Reported: repeated fresh-session questions across models; comparison of answer consistency", "No held-out evaluation established for this reported comparison" ], "missing_details": [ "Reported agreement across 15 defined runs is consistency, not verified correctness.", "Underlying runs and held-out artifacts unavailable.", "Wrong join cardinality can inflate totals despite correct metric meaning.", "Incorrect historical queries contaminate examples.", "Unstructured context can contain competing definitions.", "Model names in auto-transcript may be transcription artifacts; do not normalize them into verified model IDs." ], "candidate_patterns": [ "Shared semantic definitions (reported practice)", "Clarification gates (reported practice)", "Held-out validation (recommendation)" ], "cross_function_transfer_hypothesis": "Hypothesis: shared definitions of qualified lead, resolved ticket, or compliant document can stabilize AI outputs across sales, service and operations.", "transfer_constraints": [ "Domain owners must agree definitions", "Bad examples propagate errors", "Schema drift" ], "evidence_status": "creator-reported-or-static-artifact", "sponsorship_if_disclosed": "Free workshop promotes paid courses; no separate sponsor established.", "source_family": "AI Analyst Lab", "timestamps": [ "16:42 metric definitions", "22:14 joins", "26:06 verified examples", "32:51 ambiguity", "42:10 human approval" ], "accessed_on": "2026-10-18", "independently_run": false, "creator_recommendations_separate_from_reported_mechanics": [ "Specify joins/cardinality and verified examples", "Keep evaluation queries separate from runtime examples", "Review generated context", "Use narrowly triggered clarification gates" ], "reported_implementation_outside_comparison": [ "Shared YAML metric definitions", "Snowflake skills that stop for missing context, per creator statement" ], "source_sections": [ "Transcript paragraphs discussing shared metrics YAML and repeated-run comparison", "Transcript verified-query discussion: runtime examples versus held-out grading", "Transcript Snowflake ambiguity-skills discussion" ], "video_observed": false, "timestamp_confidence": "Chapter locator from stated textual source; not checked against video playback", "original_video_or_post_publication_date_independently_verified": false, "date_provenance": { "source_url": "https://aianalystlab.ai/workshops/build-the-context-that-makes-ai-data-agents-reliable/", "locator": "Workshop header; corroborated by Maven lesson header", "type": "event/workshop date, not upload date" }, "timestamp_provenance": { "source_url": "https://maven.com/p/34e8f6", "locator": "In this video chapter list", "retrieved_on": "2026-10-18", "reopened_at_utc": "2026-10-18T02:55:17Z", "confidence": "Chapter labels verified in creator lesson page; correspondence to video moments not observed", "independent_audit": "Auditor had not independently corroborated timestamps; owner subsequently reopened the specific chapter-list source" }, "source_dates": { "published": null, "updated": null, "other_date": null, "other_date_type": null, "evidence_url": "https://www.youtube.com/watch?v=LsaPQ5eQUyA", "evidence_note": "YouTube upload/update dates remain unknown. August 5 is the workshop event date in companion sources, not established upload date.", "checked_at": "2026-10-18T03:00:07Z", "status": "unknown_original_date" }, "source_date_entries": [ { "published": null, "updated": null, "other_date": null, "other_date_type": null, "evidence_url": "https://www.youtube.com/watch?v=LsaPQ5eQUyA", "evidence_note": "YouTube upload/update dates remain unknown. August 5 is the workshop event date in companion sources, not established upload date.", "checked_at": "2026-10-18T03:00:07Z", "status": "unknown_original_date" }, { "published": null, "updated": null, "other_date": "2026-08-05", "other_date_type": "workshop_event_date", "evidence_url": "https://aianalystlab.ai/workshops/build-the-context-that-makes-ai-data-agents-reliable/", "evidence_note": "Creator transcript page header identifies the August 5 workshop. No separate article publication/update date displayed.", "checked_at": "2026-10-18T03:00:07Z", "status": "verified_event_date_only" }, { "published": null, "updated": null, "other_date": "2026-08-05", "other_date_type": "workshop_event_date", "evidence_url": "https://maven.com/p/34e8f6", "evidence_note": "Maven lesson header dates the session August 5, 2026; reopened in metadata check. Chapter list verified as text only. Does not establish YouTube upload date.", "checked_at": "2026-10-18T03:00:07Z", "status": "verified_event_date_only" }, { "published": null, "updated": null, "other_date": "2026-08-05", "other_date_type": "workshop_event_date", "evidence_url": "https://aianalystlab.ai/workshops/build-the-context-that-makes-ai-data-agents-reliable/", "evidence_note": "Header states workshop event date, not article or YouTube publication date.", "checked_at": "2026-10-18T03:17:43.174030+00:00", "status": "verified_event_date_only" } ], "original_evidence_status": "Primary transcript reread. Reported small experiment, stated implementation practices, and evaluation recommendations now distinguished. Video not watched; no live behavior observed.", "evidence_interpretation": "Source documentation or reported practice only; no independent execution, effectiveness, safety, or transfer validation.", "reference_eligibility": { "eligible": true, "source_category": "firsthand_creator_build_or_practice", "reason": "Creator transcript supports a reported small comparison and shared-definition practice; training promotion is disclosed, no independent trial or causal effectiveness established.", "allowed_use": "process_reference_with_explicit_limits", "eligible_for_success_highlight": false, "effectiveness_proven": false, "independent_corroboration_status": "not_required_for_process_reference", "corroboration_links": [], "evidence_constraints": [ "Eligibility, where granted, permits a process reference only; it is not proof of usefulness, causal effectiveness, safety, or transferability.", "No workflow software executed, no video playback independently observed, and no outcomes independently measured by this audit.", "Underlying evaluation artifacts not inspected", "Consistency does not establish correctness", "Held-out-query evaluation recommended, not demonstrated in the reported comparison", "Reported agreement across 15 defined runs is consistency, not verified correctness.", "Underlying runs and held-out artifacts unavailable.", "Wrong join cardinality can inflate totals despite correct metric meaning.", "Incorrect historical queries contaminate examples.", "Unstructured context can contain competing definitions.", "Model names in auto-transcript may be transcription artifacts; do not normalize them into verified model IDs." ] }, "eligibility_register_id": "SOC02", "enrichment_applied": true, "enrichment_scope": "Additional inspected text/static artifacts; original historical record preserved in audit/input-snapshots/repo-handoff/social-builds.json.", "category": "analysis", "mechanism": "Creator reports a fresh-session retention comparison and shared YAML definitions; this session explicitly does not run the demo.", "evidence": [ { "url": "https://aianalystlab.ai/workshops/build-the-context-that-makes-ai-data-agents-reliable/", "inspection": "Relevant complete transcript discussion read, including comparison, semantic layers, examples, ambiguity and human approval.", "locators": [ "lines 37–54: no live demo; 15 undefined versus 15 defined runs reported", "lines 70–86: metric specification and cardinality", "lines 103–119: examples and narrow gates", "lines 130–149: synthesis into reviewed definitions" ] } ], "limitations_and_failure_cases": [ "Reported agreement across 15 defined runs is consistency, not verified correctness.", "Underlying runs and held-out artifacts unavailable.", "Wrong join cardinality can inflate totals despite correct metric meaning.", "Incorrect historical queries contaminate examples.", "Unstructured context can contain competing definitions.", "Model names in auto-transcript may be transcription artifacts; do not normalize them into verified model IDs." ], "reusable_pattern_interpretation": "Interpretation: the transferable job is answer a consequential business question under an agreed meaning. Sales qualification, service resolution and operations reporting all need domain-owned definitions plus evidence checking.", "workflow_granularity": "Definition authoring is an upstream maintenance workflow. Answering a question is the runtime workflow. Evaluation is a separate improvement loop; do not collapse all three into a single prompt.", "creator_guidance_steps_not_observed_execution": [ "Read shared definition before analysis.", "Use table relationships/cardinality to avoid duplicated rows.", "Use checked question-query examples; reserve separate examples for evaluation.", "Stop on meaning-changing ambiguity, rather than gating every question.", "Draft definitions from unstructured material, then obtain domain-owner review." ], "creator_guidance_steps_basis": "Mixed reported practices and recommendations from transcript; held-out evaluation is recommended, not demonstrated. The steps field retains only the reported fresh-session comparison." }, { "id": "SOC03", "title": "LLM Wiki", "creator": "Andrej Karpathy", "platform": "X + GitHub gist", "original_url": "https://x.com/karpathy/status/2039805659525644595", "companion_repo_or_walkthrough": [ "https://gist.github.com/karpathy/442a6bf555914893e9891c11519de94f" ], "observed_or_claimed_mechanics": "Author describes working beside Obsidian and supplies an abstract operating pattern, explicitly not a specific implementation.", "inputs": [ "Curated immutable sources", "Questions", "Schema conventions" ], "outputs": [ "Linked Markdown knowledge pages", "Cited synthesis", "Index and chronological log" ], "steps": [ "Ingest one source with human review; update summary, related pages, index and log.", "Query index first, then relevant pages; retain useful answers as new pages.", "Lint contradictions, superseded claims, orphan pages and missing links." ], "quality_checks_known": [ "Author recommends human review, source preservation, contradiction checks and version history; operation and correctness not independently observed." ], "missing_details": [ "Ingest/query/lint are proposed workflows, not a benchmarked packaged system.", "Raw sources and generated synthesis are separate layers.", "Local images require separate inspection from Markdown text.", "Author claims of negligible maintenance cost or reliable updating lack measured support.", "X access/date remains unverified." ], "candidate_patterns": [ "Compile knowledge once, maintain over time", "Separate raw evidence from synthesis" ], "cross_function_transfer_hypothesis": "Hypothesis: build a maintained customer-research or policy knowledge layer with citations and explicit supersession, reusable across onboarding, planning and support.", "transfer_constraints": [ "Permission boundaries", "Contamination from mistaken synthesis", "Source freshness", "Human adjudication" ], "evidence_status": "creator-reported-or-static-artifact", "sponsorship_if_disclosed": "No sponsorship disclosed in gist.", "source_family": "Karpathy LLM Wiki", "timestamps": [], "accessed_on": "2026-10-18", "independently_run": false, "video_observed": false, "timestamp_confidence": "Not applicable: no timestamps retained", "original_video_or_post_publication_date_independently_verified": false, "date_provenance": { "source_url": "https://gist.github.com/karpathy/442a6bf555914893e9891c11519de94f", "locator": "Gist header creation timestamp; original body only, not third-party comments", "type": "gist creation, not X publication" }, "source_dates": { "published": null, "updated": null, "other_date": null, "other_date_type": null, "evidence_url": "https://x.com/karpathy/status/2039805659525644595", "evidence_note": "Original X fetch returned 403. Secondary sources disagree about post timing; no original X publication/update date established.", "checked_at": "2026-10-18T03:00:07Z", "status": "unknown_original_date" }, "source_date_entries": [ { "published": null, "updated": null, "other_date": null, "other_date_type": null, "evidence_url": "https://x.com/karpathy/status/2039805659525644595", "evidence_note": "Original X fetch returned 403. Secondary sources disagree about post timing; no original X publication/update date established.", "checked_at": "2026-10-18T03:00:07Z", "status": "unknown_original_date" }, { "published": null, "updated": null, "other_date": "2026-04-04T16:25", "other_date_type": "gist_creation_timestamp_timezone_unspecified", "evidence_url": "https://gist.github.com/karpathy/442a6bf555914893e9891c11519de94f", "evidence_note": "Original gist header explicitly gives creation timestamp. No last-update date established. Comments are separate third-party artifacts.", "checked_at": "2026-10-18T03:00:07Z", "status": "verified_gist_creation_only" }, { "published": null, "updated": null, "other_date": "2026-04-04T16:25", "other_date_type": "gist_creation_timestamp_timezone_unspecified", "evidence_url": "https://gist.github.com/karpathy/442a6bf555914893e9891c11519de94f", "evidence_note": "Gist header creation timestamp; timezone unspecified. Original X date not inferred.", "checked_at": "2026-10-18T03:17:43.174030+00:00", "status": "verified_gist_creation_only" } ], "original_evidence_status": "Primary gist read. X fetch returned 403; X URL traced from secondary references. Treat gist as authoritative mechanics source.", "evidence_interpretation": "Source documentation or reported practice only; no independent execution, effectiveness, safety, or transfer validation.", "reference_eligibility": { "eligible": true, "source_category": "firsthand_creator_build_or_practice", "reason": "Creator’s original gist supports an abstract operating pattern and author-reported use; not a tested packaged build or proof of effectiveness.", "allowed_use": "process_reference_with_explicit_limits", "eligible_for_success_highlight": false, "effectiveness_proven": false, "independent_corroboration_status": "not_required_for_process_reference", "corroboration_links": [], "evidence_constraints": [ "Eligibility, where granted, permits a process reference only; it is not proof of usefulness, causal effectiveness, safety, or transferability.", "No workflow software executed, no video playback independently observed, and no outcomes independently measured by this audit.", "No benchmark supplied", "No guaranteed completeness or update correctness", "Ingest/query/lint are proposed workflows, not a benchmarked packaged system.", "Raw sources and generated synthesis are separate layers.", "Local images require separate inspection from Markdown text.", "Author claims of negligible maintenance cost or reliable updating lack measured support.", "X access/date remains unverified." ] }, "eligibility_register_id": "SOC03", "enrichment_applied": true, "enrichment_scope": "Additional inspected text/static artifacts; original historical record preserved in audit/input-snapshots/repo-handoff/social-builds.json.", "category": "knowledge_maintenance", "mechanism": "Author describes working beside Obsidian and supplies an abstract operating pattern, explicitly not a specific implementation.", "evidence": [ { "url": "https://gist.github.com/karpathy/442a6bf555914893e9891c11519de94f", "inspection": "Full original gist body read; third-party comments excluded.", "locators": [ "Architecture", "Operations", "Indexing and logging", "Tips and tricks", "Note" ] } ], "limitations_and_failure_cases": [ "Ingest/query/lint are proposed workflows, not a benchmarked packaged system.", "Raw sources and generated synthesis are separate layers.", "Local images require separate inspection from Markdown text.", "Author claims of negligible maintenance cost or reliable updating lack measured support.", "X access/date remains unverified." ], "reusable_pattern_interpretation": "Interpretation: a sales-account dossier or policy knowledge base can reuse this lifecycle, but requires explicit supersession and permission handling. Index and log serve different needs: navigation versus change history.", "workflow_granularity": "Three distinct jobs: incorporate evidence, answer a question, repair accumulated knowledge. A single query should not silently authorize uncontrolled changes to shared reference material.", "steps_basis": "Author-proposed operating pattern, not a verified packaged implementation." }, { "id": "SOC06", "title": "The Future of RAG is Agentic", "creator": "Cole Medin", "platform": "YouTube + creator GitHub", "original_url": "https://www.youtube.com/watch?v=_R-ff4ZMLC8", "companion_repo_or_walkthrough": [ "https://github.com/coleam00/cole-medin-knowledge-base/blob/main/sources/the-future-of-rag-is-agentic-learn-this-strategy-now.md", "https://github.com/coleam00/ottomator-agents", "https://raw.githubusercontent.com/coleam00/cole-medin-knowledge-base/main/raw/the-future-of-rag-is-agentic-learn-this-strategy-now.md" ], "observed_or_claimed_mechanics": "Timestamped creator transcript describes expanding retrieval from ranked chunks to page listing and complete-page reconstruction.", "inputs": [ "Documentation sitemap", "User question" ], "outputs": [ "Stored chunks with metadata", "Answer using retrieved evidence" ], "steps": [ "Crawl pages, chunk respecting structural boundaries, generate titles/summaries and embeddings.", "Retrieve five matching chunks filtered by source.", "If insufficient, list candidate URLs and retrieve all chunks for the chosen page.", "Combine chunks into page content; retain tool messages in conversation history." ], "quality_checks_known": [ "Creator reports an illustrative pretested comparison and acknowledges alternative retrieval fixes; no held-out benchmark or independent execution." ], "missing_details": [ "Illustrative pretested example, not a held-out benchmark.", "Creator acknowledges chunking/retrieval changes might also fix the example.", "Zero-vector fallback keeps processing moving but can degrade retrieval.", "No independent success-rate or speed evidence.", "Page fetch reconstructs stored content; it is not necessarily a fresh website fetch." ], "candidate_patterns": [ "Escalate retrieval granularity", "Let retrieval respond to evidence gaps" ], "cross_function_transfer_hypothesis": "Hypothesis: employee-policy or customer-support assistants could widen from snippet to section/full document when exceptions or context are needed.", "transfer_constraints": [ "Access controls must follow retrieval", "Context limits", "Citation fidelity", "Bound search loops" ], "evidence_status": "creator-reported-or-static-artifact", "sponsorship_if_disclosed": "GPUStack sponsorship explicitly disclosed at transcript 21:36; retain separately from recommendations.", "source_family": "Cole Medin", "timestamps": [ "33:57 basic query", "34:45 incomplete retrieval", "37:17 extra tools", "40:36 retest, per creator summary" ], "accessed_on": "2026-10-18", "independently_run": false, "video_observed": false, "timestamp_confidence": "Chapter locator from stated textual source; not checked against video playback", "original_video_or_post_publication_date_independently_verified": false, "date_provenance": { "source_url": "https://github.com/coleam00/cole-medin-knowledge-base/blob/main/sources/the-future-of-rag-is-agentic-learn-this-strategy-now.md", "locator": "Creator-owned summary frontmatter published field; secondary Glasp index differs by one day", "type": "summary-reported video date; actual upload date unresolved" }, "timestamp_provenance": { "source_url": "https://github.com/coleam00/cole-medin-knowledge-base/blob/main/sources/the-future-of-rag-is-agentic-learn-this-strategy-now.md", "locator": "Key moments in creator-owned summary", "confidence": "Summary-reported; not playback verified" }, "source_dates": { "published": null, "updated": null, "other_date": "2025-01-15", "other_date_type": "creator_summary_reported_video_publication_date_disputed", "evidence_url": "https://www.youtube.com/watch?v=_R-ff4ZMLC8", "evidence_note": "Creator-owned summary reports January 15; secondary Glasp index reports January 16. Direct original-video publication metadata was not established, so published remains null.", "checked_at": "2026-10-18T03:00:07Z", "status": "unresolved_conflicting_reported_dates" }, "source_date_entries": [ { "published": null, "updated": null, "other_date": "2025-01-15", "other_date_type": "creator_summary_reported_video_publication_date_disputed", "evidence_url": "https://www.youtube.com/watch?v=_R-ff4ZMLC8", "evidence_note": "Creator-owned summary reports January 15; secondary Glasp index reports January 16. Direct original-video publication metadata was not established, so published remains null.", "checked_at": "2026-10-18T03:00:07Z", "status": "unresolved_conflicting_reported_dates" }, { "published": null, "updated": "2026-07-21", "other_date": "2026-07-21", "other_date_type": "summary_frontmatter_created_date", "evidence_url": "https://github.com/coleam00/cole-medin-knowledge-base/blob/main/sources/the-future-of-rag-is-agentic-learn-this-strategy-now.md", "evidence_note": "Creator-owned GitHub summary frontmatter states created and updated July 21, 2026. Its published field refers to the video (January 15), not this summary. These are author-provided frontmatter dates, not independently established Git commit dates.", "checked_at": "2026-10-18T03:00:07Z", "status": "author_metadata_summary_dates" }, { "published": null, "updated": null, "other_date": null, "other_date_type": null, "evidence_url": "https://github.com/coleam00/ottomator-agents", "evidence_note": "No publication/update date established for this companion source; no inference from search crawl dates or relative age.", "checked_at": "2026-10-18T03:00:07Z", "status": "unknown" }, { "published": "2025-01-15", "updated": null, "other_date": "2026-07-21", "other_date_type": "transcript_fetched_at", "evidence_url": "https://raw.githubusercontent.com/coleam00/cole-medin-knowledge-base/main/raw/the-future-of-rag-is-agentic-learn-this-strategy-now.md", "evidence_note": "Creator-hosted transcript front matter reports source video published date; not independently checked against YouTube. Fetched timestamp belongs to transcript ingestion.", "checked_at": "2026-10-18T03:17:43.174030+00:00", "status": "creator_transcript_reports_original_date" } ], "original_evidence_status": "Creator-owned GitHub summary read; video not watched. Older foundational example, not a new 2026 build.", "evidence_interpretation": "Source documentation or reported practice only; no independent execution, effectiveness, safety, or transfer validation.", "reference_eligibility": { "eligible": true, "source_category": "firsthand_creator_build_or_practice", "reason": "Creator-owned transcript supports retrieval mechanics and reported failure/retest. GPUStack sponsorship is an isolated ad at 21:36–23:32; demonstrated stack uses OpenAI and Supabase, not GPUStack. Retain mechanism evidence, not sponsor or performance claims.", "allowed_use": "process_reference_with_explicit_limits", "eligible_for_success_highlight": false, "effectiveness_proven": false, "independent_corroboration_status": "not_required_for_process_reference", "corroboration_links": [], "evidence_constraints": [ "Eligibility, where granted, permits a process reference only; it is not proof of usefulness, causal effectiveness, safety, or transferability.", "No workflow software executed, no video playback independently observed, and no outcomes independently measured by this audit.", "Creator-hosted timestamped raw transcript was recovered and read; video playback and independent execution were not performed.", "No general accuracy benchmark reviewed", "Illustrative pretested example, not a held-out benchmark.", "Creator acknowledges chunking/retrieval changes might also fix the example.", "Zero-vector fallback keeps processing moving but can degrade retrieval.", "No independent success-rate or speed evidence.", "Page fetch reconstructs stored content; it is not necessarily a fresh website fetch." ] }, "eligibility_register_id": "SOC06", "enrichment_applied": true, "enrichment_scope": "Additional inspected text/static artifacts; original historical record preserved in audit/input-snapshots/repo-handoff/social-builds.json.", "category": "answering_questions", "mechanism": "Timestamped creator transcript describes expanding retrieval from ranked chunks to page listing and complete-page reconstruction.", "evidence": [ { "url": "https://raw.githubusercontent.com/coleam00/cole-medin-knowledge-base/main/raw/the-future-of-rag-is-agentic-learn-this-strategy-now.md", "inspection": "Complete relevant timestamped walkthrough read as transcript; no video playback.", "locators": [ "19:08: summaries use first 1,000 characters", "20:09: embedding-error fallback returns zero vector", "32:20: five chunks and source filter", "34:45–35:47: incomplete weather example", "36:33–37:03: acknowledges alternative RAG improvements", "39:22–41:21: whole-page retrieval and reported retest", "21:36: GPUStack sponsorship disclosure" ] } ], "limitations_and_failure_cases": [ "Illustrative pretested example, not a held-out benchmark.", "Creator acknowledges chunking/retrieval changes might also fix the example.", "Zero-vector fallback keeps processing moving but can degrade retrieval.", "No independent success-rate or speed evidence.", "Page fetch reconstructs stored content; it is not necessarily a fresh website fetch." ], "reusable_pattern_interpretation": "Interpretation: support and policy assistants can escalate from a paragraph hit to a complete procedure before answering. Retrieval completeness and freshness are separate checks.", "workflow_granularity": "Separate corpus preparation from question answering. Search, inspect and broaden retrieval are steps inside answering; switching databases is an implementation choice, not the reusable work pattern.", "steps_basis": "Creator transcript walkthrough, not independently run." }, { "id": "SOC07", "title": "Meeting-to-task agent with Airtable and Fireflies", "creator": "Mark Shcherbakov", "platform": "YouTube + n8n creator template", "original_url": "https://www.youtube.com/watch?v=0TyX7G00x3A", "companion_repo_or_walkthrough": [ "https://n8n.io/workflows/2683-ai-agent-for-project-management-and-meetings-with-airtable-and-fireflies/" ], "observed_or_claimed_mechanics": "Creator walkthrough describes a webhook-triggered chain that changes task and communication systems.", "inputs": [ "Completed Fireflies meeting event", "Transcript and participant details" ], "outputs": [ "Airtable tasks", "Client-specific email notifications", "Conditional Google Calendar follow-up events" ], "steps": [ "Receive completion webhook.", "Retrieve transcript and participants via authenticated GraphQL.", "Pass context and system instructions to chat model.", "Create structured Airtable records.", "Notify clients by email; create follow-up events when needed." ], "quality_checks_known": [ "No approval or duplicate-event controls established in inspected creator guide; task schema and recipient/date inference remain unresolved." ], "missing_details": [ "No approval or duplicate-event controls documented in guide.", "Exact task schema, recipient resolution and date inference not established from primary page.", "No measured meeting outcome or execution verified.", "Underlying primary JSON unavailable in inspected page; third-party copies not treated as creator-verified code.", "Notification and calendar actions are consequential outputs, not merely summaries." ], "candidate_patterns": [ "Conversation to structured commitment", "Event-triggered workflow" ], "cross_function_transfer_hypothesis": "Hypothesis: convert interviews, field visits or incident reviews into reviewable work queues while preserving source spans for each commitment.", "transfer_constraints": [ "Recording consent", "Do not invent owners or dates", "Human review before communications", "Idempotency" ], "evidence_status": "creator-reported-or-static-artifact", "sponsorship_if_disclosed": "Vendor-hosted community template; no separate sponsorship disclosed.", "source_family": "Mark Shcherbakov", "timestamps": [], "accessed_on": "2026-10-18", "independently_run": false, "video_observed": false, "timestamp_confidence": "Not applicable: no timestamps retained", "original_video_or_post_publication_date_independently_verified": false, "date_provenance": { "type": "unknown", "locator": "Template shows relative update age only; no exact original publication date established" }, "source_dates": { "published": null, "updated": null, "other_date": null, "other_date_type": null, "evidence_url": "https://www.youtube.com/watch?v=0TyX7G00x3A", "evidence_note": "Exact original video publication/update dates unavailable. Final exact-ID/date search returned no results. Companion relative update age is deliberately not converted to a date.", "checked_at": "2026-10-18T03:00:07Z", "status": "unknown_original_date" }, "source_date_entries": [ { "published": null, "updated": null, "other_date": null, "other_date_type": null, "evidence_url": "https://www.youtube.com/watch?v=0TyX7G00x3A", "evidence_note": "Exact original video publication/update dates unavailable. Final exact-ID/date search returned no results. Companion relative update age is deliberately not converted to a date.", "checked_at": "2026-10-18T03:00:07Z", "status": "unknown_original_date" }, { "published": null, "updated": null, "other_date": null, "other_date_type": null, "evidence_url": "https://n8n.io/workflows/2683-ai-agent-for-project-management-and-meetings-with-airtable-and-fireflies/", "evidence_note": "Creator template has relative last-update age only. No exact publication/update date established; no date inferred from relative age.", "checked_at": "2026-10-18T03:00:07Z", "status": "unknown_exact_dates_relative_age_excluded" }, { "published": null, "updated": null, "other_date": "8 months ago", "other_date_type": "relative_last_update_label", "evidence_url": "https://n8n.io/workflows/2683-ai-agent-for-project-management-and-meetings-with-airtable-and-fireflies/", "evidence_note": "Page displays relative Last update 8 months ago; do not convert rounded relative label into exact date.", "checked_at": "2026-10-18T03:17:43.174030+00:00", "status": "relative_label_only" } ], "original_evidence_status": "Primary creator template read; video link verified; video not watched.", "evidence_interpretation": "Source documentation or reported practice only; no independent execution, effectiveness, safety, or transfer validation.", "reference_eligibility": { "eligible": true, "source_category": "firsthand_creator_build_or_practice", "reason": "Named creator’s public n8n community template supports specified automation mechanics; vendor hosting does not by itself make it a vendor-authored customer success story. No run or usefulness verification.", "allowed_use": "process_reference_with_explicit_limits", "eligible_for_success_highlight": false, "effectiveness_proven": false, "independent_corroboration_status": "not_required_for_process_reference", "corroboration_links": [], "evidence_constraints": [ "Eligibility, where granted, permits a process reference only; it is not proof of usefulness, causal effectiveness, safety, or transferability.", "No workflow software executed, no video playback independently observed, and no outcomes independently measured by this audit.", "Only relative update date shown", "Workflow JSON not executed", "Owner/date ambiguity, duplicate events, retries and consent not specified", "No approval or duplicate-event controls documented in guide.", "Exact task schema, recipient resolution and date inference not established from primary page.", "No measured meeting outcome or execution verified.", "Underlying primary JSON unavailable in inspected page; third-party copies not treated as creator-verified code.", "Notification and calendar actions are consequential outputs, not merely summaries." ] }, "eligibility_register_id": "SOC07", "enrichment_applied": true, "enrichment_scope": "Additional inspected text/static artifacts; original historical record preserved in audit/input-snapshots/repo-handoff/social-builds.json.", "category": "meeting_follow_through", "mechanism": "Creator walkthrough describes a webhook-triggered chain that changes task and communication systems.", "evidence": [ { "url": "https://n8n.io/workflows/2683-ai-agent-for-project-management-and-meetings-with-airtable-and-fireflies/", "inspection": "Entire creator-authored guide and setup section read.", "locators": [ "Video Guide", "What this workflow does", "Setup: steps 1–6" ] } ], "limitations_and_failure_cases": [ "No approval or duplicate-event controls documented in guide.", "Exact task schema, recipient resolution and date inference not established from primary page.", "No measured meeting outcome or execution verified.", "Underlying primary JSON unavailable in inspected page; third-party copies not treated as creator-verified code.", "Notification and calendar actions are consequential outputs, not merely summaries." ], "reusable_pattern_interpretation": "Interpretation: delivery handoffs and account follow-ups share the extraction-to-tracking pattern. Add an explicit review checkpoint for owner, due date, recipient and meeting intent; track processed meeting IDs to avoid duplicates. These are proposed safeguards, not observed features.", "workflow_granularity": "Capture and extraction support the user job: turn discussion into owned commitments. Task creation, emailing and scheduling should be explicit gated branches, not hidden side effects.", "steps_basis": "Creator-authored setup/walkthrough, not independently run." }, { "id": "SOC09", "title": "Build an AI Analyst with Claude Code", "creator": "Sumeet Marwaha, interviewed by Peter Yang", "platform": "YouTube + creator article", "original_url": "https://www.youtube.com/watch?v=L3vmxY7mkNQ", "companion_repo_or_walkthrough": [ "https://creatoreconomy.so/p/build-an-ai-data-analyst-with-claude-code-sumeet" ], "observed_or_claimed_mechanics": "Public creator notes describe monitoring queries, investigating changes with Slack/Linear/code context, crafting findings, and estimating impact. Query-size limits and timeouts guard the context window.", "inputs": [ "Queries", "Data", "Incident and project context" ], "outputs": [ "Flagged changes", "Investigations", "Narrative", "Impact estimates" ], "steps": [ "Monitor", "Explore", "Craft", "Estimate impact" ], "quality_checks_known": [ "Row limits", "Query timeouts and rewrites", "No independent accuracy results inspected" ], "missing_details": [ "Article partly paywalled", "Video not watched", "End-to-end autonomous intervention not established" ], "candidate_patterns": [ "Combine quantitative anomaly with operational context", "Bound tool output" ], "cross_function_transfer_hypothesis": "Hypothesis: HR operations, supply-chain or support analysts can distinguish real behavioral changes from instrumentation and process incidents before recommending action.", "transfer_constraints": [ "Semantic definitions", "Authorized data joins", "Causal claims require stronger evidence" ], "evidence_status": "creator-reported-or-static-artifact", "sponsorship_if_disclosed": "Creator article explicitly sponsored by Glue.", "source_family": "Peter Yang / Behind the Craft", "timestamps": [ "03:04 framework", "10:23 MCP demo", "21:10 context", "26:04 Slack/Drive" ], "accessed_on": "2026-10-18", "independently_run": false, "video_observed": false, "timestamp_confidence": "Chapter locator from stated textual source; not checked against video playback", "original_video_or_post_publication_date_independently_verified": false, "date_provenance": { "type": "creator_companion_article_date", "locator": "Public creator article byline; not independently verified video upload date" }, "source_dates": { "published": null, "updated": null, "other_date": null, "other_date_type": null, "evidence_url": "https://www.youtube.com/watch?v=L3vmxY7mkNQ", "evidence_note": "Original YouTube identity verified through creator link, but accessible video page did not expose upload/update metadata. January 18 dates creator companion article.", "checked_at": "2026-10-18T03:00:07Z", "status": "unknown_original_date" }, "source_date_entries": [ { "published": null, "updated": null, "other_date": null, "other_date_type": null, "evidence_url": "https://www.youtube.com/watch?v=L3vmxY7mkNQ", "evidence_note": "Original YouTube identity verified through creator link, but accessible video page did not expose upload/update metadata. January 18 dates creator companion article.", "checked_at": "2026-10-18T03:00:07Z", "status": "unknown_original_date" }, { "published": "2026-01-18", "updated": null, "other_date": null, "other_date_type": null, "evidence_url": "https://creatoreconomy.so/p/build-an-ai-data-analyst-with-claude-code-sumeet", "evidence_note": "Public creator article byline explicitly dates January 18, 2026. No separate update date displayed; not treated as video upload verification.", "checked_at": "2026-10-18T03:00:07Z", "status": "verified_companion_article_publication" } ], "original_evidence_status": "Primary public creator article read; YouTube identity verified. Only accessible portions used.", "evidence_interpretation": "Source documentation or reported practice only; no independent execution, effectiveness, safety, or transfer validation.", "reference_eligibility": { "eligible": true, "source_category": "firsthand_creator_build_or_practice", "reason": "Brex data practitioner describes Claude Code mechanics. Glue sponsorship is a separate ad for another product; Claude team recommended the guest, so vendor selection influence is disclosed. Retain reported mechanics only, not independent validation.", "allowed_use": "process_reference_with_explicit_limits", "eligible_for_success_highlight": false, "effectiveness_proven": false, "independent_corroboration_status": "not_required_for_process_reference", "corroboration_links": [], "evidence_constraints": [ "Eligibility, where granted, permits a process reference only; it is not proof of usefulness, causal effectiveness, safety, or transferability.", "No workflow software executed, no video playback independently observed, and no outcomes independently measured by this audit.", "Article partly paywalled", "Video not watched", "End-to-end autonomous intervention not established" ] }, "eligibility_register_id": "SOC09" }, { "id": "SOC10", "title": "Ramp product-shaping skill", "creator": "Geoff Charles / Ramp, interviewed by Peter Yang", "platform": "YouTube + creator article", "original_url": "https://www.youtube.com/watch?v=RBqT2PHWdBg", "companion_repo_or_walkthrough": [ "https://creatoreconomy.so/p/inside-ramp-the-32b-company-ai-agents-geoff-charles" ], "observed_or_claimed_mechanics": "Public notes describe problem-framing questions, parallel competitor/customer/ticket/code research, per-agent markdown findings, then conversational convergence on a short evidence-grounded specification.", "inputs": [ "Proposed idea", "Customer calls", "Support tickets", "Competitors", "Code" ], "outputs": [ "Research findings", "Specification with alternatives and open questions" ], "steps": [ "Challenge framing", "Research in parallel", "Synthesize patterns", "Shape specification" ], "quality_checks_known": [ "Pushback on unsupported framing", "Evidence-grounded requirements", "Alternatives and unresolved questions retained" ], "missing_details": [ "Full skill paywalled", "Speed and adoption claims not validated", "No research scoring rubric inspected" ], "candidate_patterns": [ "Frame before researching", "Parallel evidence gathering with shared synthesis", "Preserve alternatives" ], "cross_function_transfer_hypothesis": "Hypothesis: procurement, operations and enablement can reuse the framing-to-evidence-to-brief sequence, substituting their own sources and decision criteria.", "transfer_constraints": [ "Parallel agents can repeat errors", "Sources need deduplication", "Decision owner remains accountable" ], "evidence_status": "creator-reported-or-static-artifact", "sponsorship_if_disclosed": "Public article discloses Granola partnership; guest is describing own employer.", "source_family": "Peter Yang / Behind the Craft", "timestamps": [ "04:45 customer agent", "07:57 analyst", "13:08 shaping skill" ], "accessed_on": "2026-10-18", "independently_run": false, "video_observed": false, "timestamp_confidence": "Chapter locator from stated textual source; not checked against video playback", "original_video_or_post_publication_date_independently_verified": false, "date_provenance": { "type": "creator_companion_article_date", "locator": "Public creator article byline; not independently verified video upload date" }, "source_dates": { "published": null, "updated": null, "other_date": null, "other_date_type": null, "evidence_url": "https://www.youtube.com/watch?v=RBqT2PHWdBg", "evidence_note": "Original YouTube identity verified through creator link, but accessible video page did not expose upload/update metadata. March 15 dates creator companion article.", "checked_at": "2026-10-18T03:00:07Z", "status": "unknown_original_date" }, "source_date_entries": [ { "published": null, "updated": null, "other_date": null, "other_date_type": null, "evidence_url": "https://www.youtube.com/watch?v=RBqT2PHWdBg", "evidence_note": "Original YouTube identity verified through creator link, but accessible video page did not expose upload/update metadata. March 15 dates creator companion article.", "checked_at": "2026-10-18T03:00:07Z", "status": "unknown_original_date" }, { "published": "2026-03-15", "updated": null, "other_date": null, "other_date_type": null, "evidence_url": "https://creatoreconomy.so/p/inside-ramp-the-32b-company-ai-agents-geoff-charles", "evidence_note": "Public creator article header explicitly dates March 15, 2026. No separate update date displayed; not treated as video upload verification.", "checked_at": "2026-10-18T03:00:07Z", "status": "verified_companion_article_publication" } ], "original_evidence_status": "Primary public creator article read; linked video identity verified; video not watched.", "evidence_interpretation": "Source documentation or reported practice only; no independent execution, effectiveness, safety, or transfer validation.", "reference_eligibility": { "eligible": true, "source_category": "firsthand_creator_build_or_practice", "reason": "Ramp CPO describes an internal Claude Code product-shaping skill. Granola partnership is a separate ad; the retained skill is not a Granola product showcase. Retain public mechanics with sponsorship disclosure.", "allowed_use": "process_reference_with_explicit_limits", "eligible_for_success_highlight": false, "effectiveness_proven": false, "independent_corroboration_status": "not_required_for_process_reference", "corroboration_links": [], "evidence_constraints": [ "Eligibility, where granted, permits a process reference only; it is not proof of usefulness, causal effectiveness, safety, or transferability.", "No workflow software executed, no video playback independently observed, and no outcomes independently measured by this audit.", "Full skill paywalled", "Speed and adoption claims not validated", "No research scoring rubric inspected" ] }, "eligibility_register_id": "SOC10" } ]