Meta Ad Library, Google Ads Transparency & Video Ad Analyzer
Discussions
From an advertiser search to a source-linked review shortlist For readers who have not yet chosen a video to analyze, our existing Google Ads Transparency walkthrough follows a manual canva.com search into two review candidates: https://github.com/LMSlay/tubetext-examples/blob/main/recipes/google-ads-canva-source-inventory.md The observations are dated 24 October 2026, not a fresh check of Canva's ads. The example separates the selected advertiser identity from a domain search, records the selected US display region and filters, and retains each creative's source link. Only the first preview variant was inspected. Visible preview copy is not a verified speech transcript; two creative IDs do not prove two different video assets. Before using the example for current work, reopen the permitted original sources and confirm the region, variant and playable media. Keep unresolved entries as metadata-only records rather than inventing shots or narration. The guide's ledger is a manual research record, not an Actor output schema, and its input example was not executed. This is a discovery and source-selection resource. A last-shown date or preview does not establish spend, conversions, ROAS or a winning creative. TubeText is not affiliated with Canva or Google. Prepared by TubeText Labs with AI assistance; collection is optional, subject to current Store pricing and source rights. No new Actor run or MCP integration test was performed.
Reviewing an existing Shot-by-Shot export If you already have a Shot-by-Shot Video Analysis Pro export, our review guide starts with that retained file: https://github.com/LMSlay/tubetext-examples/blob/main/recipes/shot-export-rows-and-overlay-evidence.md Separate shot rows, whole-video context and unresolved row types before counting. Keep the original timing and descriptions alongside any manual corrections. A retained-row count is not proof of complete video coverage. If an audience overlay is present, preserve its actual association label: timestamp, text_mention, semantic or none. A semantic link should not become “viewers reacted to this shot” in a handoff. Missing overlay data does not establish zero reaction, and an association does not prove retention or sales impact. The guide includes an invented review worksheet, not collected output or a live test. It concerns the Shot-by-Shot product listed in this connector, not the separate Video Breakdown AI n8n example. Reading it needs no new run; Actor use is optional and billed separately under current Store pricing and applicable Apify charges. Published by TubeText Labs with AI assistance. No Actor or MCP integration test was run for this resource; source-rights and human review still apply.
Glama MCP Gateway
Add one secure layer between your agents and this server.
Maintainer resource note (TubeText; AI-assisted): When comparing competitor ads, keep the source observation separate from the AI interpretation and from your own creative hypothesis. Our existing guide uses four review blocks: source observation, AI interpretation, your original hypothesis, and review & measurement. Use it to explain your choice of angle, not to label an ad a winner. Guide: https://tubetext.dev/video-creative-research/#brief This is a documentation-based editorial template, not an Actor output schema, a live test, or a measured campaign result. The guide is free to read without running this connector. Optional TubeText/Apify use is billed separately under the relevant Actor’s current pricing. It does not establish spend, conversions, ROAS, or permission to reuse another advertiser’s assets. Human review and source-rights checks remain necessary.