Tools reference

Use it whenRead your channel's views, watch hours, subscribers, traffic and top uploads, with live counts and the last 48 hours.

You provide

A connected channel and a 7, 28 or 90-day period.

You receive

Reported totals with their last reported day, live counts not yet reported, and daily and hourly views.

Use it whenSee which uploads stand out on click-through rate, retention, subscriber conversion and engagement against your own norms.

You provide

A connected channel.

You receive

Each recent upload's results against its same-type peers (click-through rate for uploads from the last 28 days), with its clearest strength and weakness.

Use it whenRead one upload's reach, retention drop-offs, engagement, search terms, recommending videos and audience.

You provide

One of your uploads.

You receive

Results since publishing, impressions and click-through rate by source, retention moments and audience breakdowns.

Use it whenCheck current AI usage, remaining budget, and whether a planned operation can run.

You provide

Optionally name the AI operation you are considering and provide your own cost estimate.

You receive

Your plan, billing window, spent and reserved amounts, remaining budget, estimated operation cost, and whether it can run.

Use it whenEstimate AI cost before running a title and thumbnail packaging workflow.

You provide

Choose whether to include title variants, a thumbnail prompt, thumbnail images, generation quality, and the number of images.

You receive

A total estimated cost, a line-item breakdown, and your budget status for the proposed workflow.

Use it whenRead the signed-in creator's saved channel context for personalized creative work.

You provide

Usually nothing. The primary or latest channel is used automatically, but you can specify one of your channel IDs.

You receive

Your saved creative blueprint and direction health, with extracted channel context separately when available. Request a previous version to review or roll back.

Use it whenSave a confirmed change to your channel's creative direction with revision history.

You provide

Current version and the fields to change. Preview first, then confirm the same patch with the returned confirmation token and idempotency key.

You receive

Updated blueprint version and direction health. Previous versions remain readable for rollback. Creator memory refreshes from saved inputs on its next read.

Use it whenGather channel context, relevant techniques, and title-format candidates for a topic.

You provide

An optional topic and description, plus optional channel and result limits.

You receive

Your relevant channel blueprint, ideation techniques, title formats, creative constraints, and a suggested next-step flow.

Use it whenGenerate structured video angles using Mittalmar techniques and channel context without saving them.

You provide

A topic, with optional context, preferred techniques, channel, idea count, or source-video inspiration.

You receive

Draft ideas with descriptions, the technique used, and fit scores covering channel fit, clarity, curiosity, and reach.

Use it whenEvaluate a draft idea for channel fit, clarity, curiosity, and potential reach.

You provide

The idea to evaluate, plus an optional description or channel.

You receive

An overall score, dimension scores, strengths, risks, and the recommended next tool.

Use it whenSelect distinct title structures that can package different angles for one video.

You provide

A video topic, with optional Project, description, technique, origin topic, or channel. A Project supplies its current selected research direction.

You receive

Recommended title formats with IDs, descriptions, relevance scores, and reasons for each recommendation.

Use it whenGenerate structured title options from selected or recommended formats without saving them.

You provide

A topic, with optional Project, description, format choices, inspiration titles, technique, channel, and number of variants. A Project automatically supplies its current explicitly selected research angle.

You receive

Title variants with their format, clarity, curiosity, length and overall fit scores, plus any risks.

Use it whenCreate title options, select a leading title, and draft a matching thumbnail prompt without saving work.

You provide

A topic or idea, plus optional Project, creative context, title-format choices, inspiration, channel, thumbnail style sources, assets, and direction. A Project supplies its current selected research direction.

You receive

Candidate titles, a selected title, a thumbnail plan and prompt, a package score, risks, and recommendations.

Use it whenList available formats, recipes, references, art styles, and assets for thumbnail planning.

You provide

Optionally choose which source types to include and how many results to return.

You receive

Available sources with IDs, types, names, descriptions, preview URLs where available, and counts by type.

Use it whenDraft a generator-ready thumbnail prompt from the video's package and available creative sources.

You provide

A video title and description, plus optional Project, style source, composition, recipe, reference, art style, assets, or creative direction. A Project automatically supplies its current selected research angle.

You receive

The selected creative source and art style, a generator-ready prompt, and practical generation notes.

Use it whenQueue confirmed thumbnail generation from a ready prompt and return a batch identifier.

You provide

An approved prompt, generation mode and image count, plus optional Project, assets, reference images, and video context. Confirmation is required; a supplied owned Project receives the generation.

You receive

A generation batch ID and status. With waitSeconds (up to 60), returns current progress and completed previews before the wait or tool deadline expires.

Use it whenCheck an owned generation batch and retrieve completed image and thumbnail URLs.

You provide

The generation batch ID; set includePreviews to false for lightweight polling.

You receive

Batch progress and image URLs, plus labeled small previews by default. Reports unavailable or truncated previews; inline display depends on your client.

Use it whenInspect one completed thumbnail from your account.

You provide

An image ID and optional size: preview (default) or full.

You receive

One image content block and its image ID and URLs. Full-size images are returned only when requested, up to 10 MiB. Display depends on your client.

Use it whenReview recently completed thumbnails from your Project.

You provide

An owned Project ID, an optional limit (1–50, default 8), and includePreviews (default true).

You receive

Recent images with URLs and optional labeled previews. Responses include at most 8 previews within a 500 KB budget and report any truncation. Use an image ID to inspect an omitted preview.

Use it whenRead a bounded page of timing data for a saved narration take.

You provide

Your owned Project and the relevant file or review settings.

You receive

A saved result, progress or evidence, with a Project link where applicable.

Use it whenBrowse default voices, your own clones, and the public voice library.

You provide

An owned script, with optional library search and page.

You receive

Voice choices, existing preview samples, and library selection tokens.

Use it whenRead script narration context and check or recover a previous generation.

You provide

An owned script and, optionally, the previous generation’s idempotency key.

You receive

Exact manuscript and opening, available models, pricing, saved takes and generation state. Recovery may finish attaching previously generated audio and its existing usage record.

Use it whenRetrieve saved narration audio and word timings.

You provide

An owned script and saved take.

You receive

Take details, timings, and an audio download link valid for 15 minutes.

Use it whenCreate a narration voice from recordings of your own voice uploaded to a Project.

You provide

An owned Project, 1–5 of its audio recordings of your voice, a voice name and your consent.

You receive

A confirmation request, followed by your new voice, ready to choose for narration.

Use it whenDelete one of your cloned voices to free a voice slot.

You provide

One of your cloned voices.

You receive

A confirmation request, followed by the deleted voice. Saved takes keep their audio.

Use it whenGenerate ElevenLabs narration using your Mittalmar AI budget.

You provide

An owned saved script, exact opening, manuscript or passage text, selected voice and narration settings.

You receive

A cost estimate and confirmation request, followed by a saved take. Reuse the same idempotency key when checking or retrying a generation.

Use it whenRead the published video linked to your Project and any changes waiting for your approval.

You provide

Your Project linked to a published upload.

You receive

The synced title, description and thumbnail, and recent proposed changes.

Use it whenPropose a new title, description, tags or thumbnail for a published video; you approve it in Mittalmar before anything changes.

You provide

Your Project, with any of a title, description, tags, a Project thumbnail and a short reason.

You receive

An approval link. Nothing is published until you approve the live before and after.

Use it whenCheck whether a proposed video change is waiting, applied, rejected or expired, and what YouTube did with it.

You provide

The change request from a proposal.

You receive

Its status and, once applied, the outcome for each part.

Use it whenApply corrected Project packaging to your private upload and clear the old review.

You provide

Your Project, with optional caption, chapter and playlist settings.

You receive

Queued package sync and the review page.

Use it whenSend a bounded file chunk for an upload you already confirmed.

You provide

Your owned Project and the relevant file or review settings.

You receive

A saved result, progress or evidence, with a Project link where applicable.

Use it whenRead file bytes in chunks when your client cannot fetch download links.

You provide

Your owned Project and the relevant file or review settings.

You receive

A saved result, progress or evidence, with a Project link where applicable.

Use it whenConfirm up to twenty editing assets, then upload and attach each file.

You provide

Your owned Project and the relevant file or review settings.

You receive

A saved result, progress or evidence, with a Project link where applicable.

Use it whenImport a public HTTPS file directly into your Project.

You provide

Your owned Project and the relevant file or review settings.

You receive

A saved result, progress or evidence, with a Project link where applicable.

Use it whenRemove a thumbnail from the Project picker, keeping its library asset.

You provide

Your owned Project and the relevant file or review settings.

You receive

A saved result, progress or evidence, with a Project link where applicable.

Use it whenCheck your connected channel and publishing permission.

You provide

Your owned Project and the relevant file or review settings.

You receive

A saved result, progress or evidence, with a Project link where applicable.

Use it whenRead private-upload progress, packaging and the Studio link.

You provide

Your owned Project and the relevant file or review settings.

You receive

A saved result, progress or evidence, with a Project link where applicable.

Use it whenUpload your finished video privately with its selected packaging and explicit policy settings.

You provide

Your owned Project and the relevant file or review settings.

You receive

A saved result, progress or evidence, with a Project link where applicable.

Use it whenCheck the private upload and original render; inspect evidence and missing checks.

You provide

Your owned Project and the relevant file or review settings.

You receive

A saved result, progress or evidence, with a Project link where applicable.

Use it whenRead the saved review report with evidence and limitations.

You provide

Your owned Project and the relevant file or review settings.

You receive

A saved result, progress or evidence, with a Project link where applicable.

Use it whenOpen the review page for a human Publish or Schedule click; never publishes unattended.

You provide

Your owned Project and the relevant file or review settings.

You receive

A saved result, progress or evidence, with a Project link where applicable.

Use it whenList compact projects owned by the signed-in customer.

You provide

Optionally provide search text and paging limits to narrow your owned projects.

You receive

Project IDs, ideas, descriptions, selected-title and image states, title and thumbnail counts, and paging metadata.

Use it whenRead one owned project with its saved titles and linked thumbnail generations.

You provide

The ID of a project owned by your signed-in account.

You receive

The Project record, linked assets, saved titles, linked generation batches, completed image URLs and optional previews, compact research decision brief with complete direction choices, selected-direction identity and paged evidence pointers. Stale or unselected research is excluded from generation context.

Use it whenCheck whether a project has enough idea, title, and thumbnail packaging to move forward.

You provide

The ID of a project owned by your signed-in account.

You receive

A package-readiness score with strengths, risks, and recommended improvements.

Use it whenLink your conversation's completed ideas, title batches and optional Script drafts to one Project for 30 days.

You provide

Choose an existing Project ID or a new Project label and optional premise. Enable Script saving when needed and confirm the Project and save scope once.

You receive

Linked session ID, Project ID and link, expiry, allowed save scope, and current revision.

Use it whenSave completed drafts from your conversation to the linked Project without another Mittalmar confirmation for each revision.

You provide

Linked session ID, a retry key, completed premise, title alternatives or Script stages with stable artifact keys, and revisions when reconciling edits. Select a Working title only when you choose one.

You receive

Created, updated, or unchanged artifact IDs and revisions, plus a saved receipt and Project link. Conflicting edits return a conflict instead of overwriting newer work.

Use it whenRead the linked Project and saved title alternatives before continuing work or reconciling edits.

You provide

Your linked session ID.

You receive

Project details, bounded title alternatives, artifact IDs, revisions, session status, and coverage.

Use it whenStop saving through this linked session while keeping existing work intact.

You provide

Your linked session ID.

You receive

Confirmation that the session has stopped and a link to the Project.

Use it whenRead an owned Script before continuing or revising it in your AI client.

You provide

An owned Project ID and Script ID.

You receive

Hook, hook context, open loops, outline, manuscript, language, revision and direct Script link.

Use it whenPrepare Mittalmar’s stage-specific writing instructions for your AI client to execute.

You provide

An owned Project, writing stage, brief, relevant source text and completed prior-stage outputs. Optionally choose an owned voice style.

You receive

Writing instructions, relevant hook, opening-style and voice context, and the Project’s current selected research direction with source coverage. This tool neither generates nor saves the draft.

Use it whenCreate up to twenty projects with a single confirmed batch.

You provide

Selected ideas with optional working titles, descriptions and structured source excerpts. Preview the batch and confirm it with one returned token and idempotency key.

You receive

Normalized previews before confirmation; project IDs and selected title IDs after saving.

Use it whenKeep attributed source excerpts with an existing project.

You provide

An owned project and sources with URLs, platform, excerpts and optional author, publication date, metrics and external IDs. Requires confirmation.

You receive

The saved source list. Matching URLs replace older entries. Saving preserves evidence without verifying its claims.

Use it whenCreate a confirmed project from an idea and optional working title.

You provide

An idea, with optional description, working title, technique, origin topic, source-video context or structured source excerpts. Preview first, then confirm with the returned token and idempotency key.

You receive

A normalized preview before confirmation; the new project ID and optional selected title ID after saving.

Use it whenSave confirmed title variants under an owned project.

You provide

A project ID and one or more titles, with optional format and origin details. You can also select the first title. Confirmation is required.

You receive

The project ID, IDs of the saved titles, and the selected title ID when one was chosen.

Use it whenBulk-create confirmed projects from selected draft ideas.

You provide

One or more selected ideas with optional descriptions, working titles, techniques, origin topics, or source-video context. Confirmation is required.

You receive

The created project IDs and any selected title IDs created from working titles.

Use it whenLink a confirmed thumbnail generation batch to an owned project.

You provide

A generation batch ID and project ID, with an optional saved title or source title. Confirmation is required.

You receive

The linked batch, project, and title IDs.

Use it whenStart attaching a file made outside Mittalmar to your Project.

You provide

Project ID, asset kind, filename, content type, byte size up to 1 GiB, and an MD5 (direct) or SHA-256 (chunks) checksum, with an optional label and thumbnail selection. Counts toward the plan's Project storage. Confirmation is required.

You receive

A ten-minute direct upload URL, or a 24-hour MCP upload with bounded chunks and checksum verification. Complete the upload after transferring the file.

Use it whenVerify and attach an uploaded file to your Project.

You provide

The Project ID and upload ID returned when starting the upload. Confirmation is required.

You receive

Asset ID, image ID for thumbnails, unlisted file URL and Project link. Retries reuse the attached asset.

Use it whenMark a confirmed completed image as an owned project's selected thumbnail.

You provide

A Project ID and a completed or uploaded thumbnail image ID, or null to clear the selection. Confirmation is required.

You receive

The project and selected image IDs, plus the selected full image and thumbnail URLs.

Use it whenSearch curated title structures and relevant examples.

You provide

Optional search text across names, descriptions, categories, and example titles; category filters, favorites-only selection, Format IDs for follow-ups, paging, and sort order.

You receive

Relevant Title Formats with matched fields, favorite state, descriptions, categories, example counts, up to three linked videos each, and paging metadata.

Use it whenSearch curated visual layouts and relevant examples.

You provide

Optional search text across names, descriptions, categories, and example titles; category filters, favorites-only selection, Format IDs for follow-ups, paging, and sort order.

You receive

Relevant Thumbnail Formats with matched fields, favorite state, descriptions, categories, example counts, up to three linked examples each, and paging metadata. Matching uses curated text, not image analysis.

Use it whenSearch reusable techniques for developing stronger video angles.

You provide

Optional search text, category filters, paging, and sort order.

You receive

Matching ideation techniques with IDs, descriptions, categories, example counts, representative videos, and paging metadata.

Use it whenSearch atomic script-opening patterns by style and creative purpose.

You provide

Optional search text, style/content/niche axes, category filters, and paging limits.

You receive

Matching hook patterns with IDs, descriptions, example wording, categories, and paging metadata.

Use it whenRead attached Project research, its progress and recommendations, alongside saved source documents.

You provide

An owned Project ID; optionally a research section (evidence, competitors, market_map, adjacent_transfers, context_citations or content_coverage), offset and limit to inspect supporting data. Asset ID, chunk position and limit (default four) read bounded saved-document passages without repeating the decision brief.

You receive

Compact decision brief, complete direction choices, current or stale state and coverage, plus explicit section read pointers, bounded supporting-data pages, Asset inventory, stored passages and next-page positions. Source evidence remains subject to its recorded limitations.

Use it whenSave research written in your AI client as a source document attached to your Project.

You provide

An owned Project ID, document name and Markdown content including source links and uncertainty. Confirmation is required.

You receive

Saved Asset ID and Project link. Different content creates a separate document; identical content is reused. Saving does not verify claims.

Use it whenFind the stored evidence available for an investigation in your AI client.

You provide

Optional source-name search, page size, independent collection and tracked-channel cursors, and includeOverview (default true).

You receive

Visible collections, favorites and explicitly tracked competitors with source IDs, all-format inventory counts and available observation dates; optional whole-scope medium/long summaries for saved and tracked evidence. Name search filters the inventory, not those summaries. Follow each source list's cursor separately; discovery does not collect new evidence.

Use it whenCompute filtered dataset comparisons for your AI client to interpret and chart.

You provide

A subject and one to four named cohorts selecting stored saved, favorites, collections or tracked competitors. Optional direction, target language, target duration and day-7/14/28 checkpoint. Cohort filters include title/channel search, source IDs, channel, publication dates and duration; the subject does not filter videos, and target language/duration qualify pattern comparisons rather than filtering descriptive totals.

You receive

Whole-scope counts, descriptive distributions, chart-ready count series, missing-data coverage, gated pattern comparisons and supporting or contrary video IDs. Stored classifications remain inferred labels; comparable performance claims are withheld when not qualified. Cohorts may overlap and are independent stored reads, not a transactional snapshot.

Use it whenBrowse the medium and long videos behind a dataset comparison.

You provide

Stored saved, favorites, collections or tracked scope; optional source IDs, search, channel, publication and duration filters, sort order, video IDs and cursor. Return up to 30 videos per page.

You receive

Deduplicated video rows with titles, original video and thumbnail links, memberships, stored metrics, observation dates, pagination and coverage. Includes only known-duration videos of at least four minutes and excludes explicit Shorts. Does not inspect image content or transcripts.

Use it whenFind your visible collections and explicitly tracked competitors, with stored evidence counts.

You provide

Optional source name or path, page size, and independent collection or tracked-channel cursors.

You receive

Collection IDs, names, paths, saved-link counts, favorites count, and tracked channel IDs with indexed-video counts and observation dates. Each source type has its own pagination; saved-link counts can include aliases.

Use it whenFind relevant videos across your favorites and collections, or stored uploads from your explicitly tracked competitors. Smaller references remain included.

You provide

Search text, saved/favorites/collections or tracked scope, collection or tracked-channel IDs, optional descendants, channel, duration, publication dates, sorting, pagination, or returned video IDs.

You receive

Deduplicated videos with titles, image and video links, selected memberships, stored metrics and freshness, match fields, and coverage. Tracked scope excludes your connected channel. Searches stored titles and channel names; never refreshes or inspects transcripts or images.

Use it whenCompute descriptive analytics across every eligible medium/long video in the selected stored evidence, without downloading a dataset.

You provide

Favorites, collections or tracked competitor scope, selected source IDs, optional descendants, search, channel, date and duration filters.

You receive

Known duration of at least four minutes with explicit Shorts excluded; exact format-exclusion coverage, deduplicated counts, missing-data denominators, publication/runtime/title-length distributions, up to 30 channel groups with remainder counts, and descriptive lifetime views with observation dates. These are not age-adjusted performance predictions.

Use it whenFind saved analysis versions for visible collections associated with your connected channel.

You provide

Optional page size and cursor.

You receive

Up to 30 stored report summaries and a next-page cursor. Reading does not refresh or process reports.

Use it whenRead a frozen collection analysis and its evidence-backed findings.

You provide

A report ID returned by collection research.

You receive

Whole-snapshot descriptive counts, classification coverage, findings and limitations. Labels are inferred from titles; frequency does not establish demand or performance.

Use it whenInspect frozen source records and title-based classifications behind a collection analysis.

You provide

A report ID, page size and cursor, or up to 20 cited video IDs.

You receive

A bounded evidence page, coverage and pagination. No transcript retrieval, image interpretation or source refresh.

Use it whenFreeze a visible owned collection for a new title and text analysis.

You provide

Collection ID, research subject and a run budget up to $3. Requires confirmation.

You receive

A report ID, coverage and scoped continuation receipt. After confirmation, bounded AI analysis continues automatically in the background within the approved cap.

Use it whenProcess one bounded AI classification or synthesis step for an existing collection analysis.

You provide

An owned report ID and optional scoped continuation receipt. A valid receipt continues the approved run; otherwise confirmation is required.

You receive

Status, coverage and actual usage within the original cap and available AI allowance. Performs a bounded step and schedules remaining analysis; a valid scoped receipt reuses the approval. Completed reports do no model work.

Use it whenCreate a separate analysis version to retry failed records while preserving the original.

You provide

An eligible report ID. Requires confirmation.

You receive

A new report version and scoped continuation receipt. Confirmed retry starts automatic background analysis within the original cap, preserving the earlier version.

Use it whenRead your current creator direction, choices and qualified observations.

You provide

Optionally supply a topic to prioritize relevant stored knowledge. A compact brief is default; use detail:full for the complete persisted snapshot.

You receive

Creator direction, relevant cited mechanisms and prior research, source dates, qualification status, explicit preview counts and deeper evidence pointers.

Use it whenRefresh your creator context from changed saved instructions and observations.

You provide

Optionally force a refresh.

You receive

Updated saved context and dated coverage; does not start paid analysis.

Use it whenInspect available age-matched video evidence and its coverage.

You provide

Select saved, collection or tracked competitor scope and follow the returned cursor.

You receive

A bounded page of medium and long videos with known duration of at least four minutes, excluding explicit Shorts, with genuine observation checkpoints, missing-history coverage and limits on comparable performance.

Use it whenCarry an angle from completed research into a Project.

You provide

A workflow ID and angle ID; requires confirmation. Project research uses its existing Project. An optional Project ID is available for older standalone research.

You receive

The existing owned Project and linked research containing the chosen premise, viewer promise, positioning and citations. A changed premise or selected research version requires refreshing before choosing. Existing creative fields are preserved.

Use it whenInspect the creator and Project source excerpts actually used in a research report.

You provide

An owned workflow ID and optional offset, limit and source hash.

You receive

Bounded original Asset passages with chunk citations, source coverage, stable hash and next page offset. Reads the frozen excerpt sample, without new acquisition or AI usage.

Use it whenReview the current draft using acquired sources, or request fresh evidence for the same Project idea.

You provide

The current owned workflow ID and a stable idempotencyKey. Optional mode: existing-sources reviews eligible acquired evidence with current creator and Script context; the default requests fresh evidence from current completed research or explicitly restarts current cancelled research.

You receive

A new background version with a cited decision brief, source dates, coverage and bounded included-allowance usage. Existing-source review retains the original acquired evidence dates; earlier versions and charges remain recorded.

Use it whenInspect acquired transcript passages and their original timestamps.

You provide

A workflow ID, source video ID, text offset and page size; optionally supply the source hash when following pages.

You receive

Bounded transcript text and timestamped excerpts, next offsets, source hash, date and explicit coverage. Missing source content remains unknown.

Use it whenResearch medium and long video topics using your creator context and included AI allowance.

You provide

A topic and optional creative direction. Research includes videos of at least four minutes with known duration, excluding Shorts and unknown-duration videos. Supply an owned Project or create a new Project from the topic before research starts; collections and tracked competitors can provide additional context. Unchanged Project premises and directions reuse current research; new Project research has no daily paid refresh schedule.

You receive

A Project Research link, workflow ID and progress. Research belongs to the Project from the start and discovers competitors and adjacent references, acquires available source content, computes analytics and proposes cited angles within one bounded allowance. Read the workflow for results.

Use it whenRead a saved topic-research workflow and its progress.

You provide

A workflow ID and optional waitSeconds from 0 to 25 (default 0). Use 25 while research is active.

You receive

Status, coverage, analytics, angles, limitations and a cost receipt, with up to 12 evidence examples. A next action guides continued monitoring of the same job; waiting starts no new research and uses no AI allowance.

Use it whenFind your latest saved topic-research workflows.

You provide

No input is required.

You receive

Metadata and retained-evidence counts for the latest 30 owned workflows. Read a workflow to retrieve its evidence examples.

Use it whenSave a new immutable research version with revised direction or sources.

You provide

A workflow ID and changes to the topic, direction, Project, collections, tracked channels, adjacent references or excluded videos. Unchanged Project premises and directions reuse current research. Set autoRefresh to false only to stop an older opted-in refresh schedule.

You receive

A new automatic research version and progress. Uses included AI allowance within a bounded cap; available source evidence can be reused. The earlier version remains available.

Use it whenResume an unfinished automatic research workflow.

You provide

A workflow ID.

You receive

Updated progress and results when available, continuing within the original bounded AI allowance. Completed and cancelled workflows remain unchanged.

Use it whenStop an unfinished workflow while keeping completed research available.

You provide

A workflow ID.

You receive

Updated workflow status; completed workflows remain unchanged.

Use it whenRead more original evidence from a saved research snapshot.

You provide

A workflow ID, offset and page size up to 50.

You receive

A bounded page of source videos with pagination and coverage. Image links are references, not proof of visual analysis.

Use it whenKeep a completed research snapshot as readable source material in a Project.

You provide

A workflow ID and owned Project ID. Requires confirmation.

You receive

A linked Markdown research Asset, content coverage and Project receipt. Includes findings, selected evidence, limitations and workflow references for further inspection. Saving does not verify claims or create a Script.

Use it whenRead current trend topics for a geography with available volume and evidence links.

You provide

Choose trend sources, a two-letter country code, an optional X location ID, and a result limit.

You receive

Ranked topics with source, volume, timing, context, supporting articles and links, plus the status of each trend source.

Use it whenUse AI to rank current trends against channel context and suggest defensible creative angles.

You provide

Choose trend sources and geography, plus an optional channel and maximum number of recommendations.

You receive

Ranked channel matches with relevance, confidence, urgency, fit reasons, risks, a content angle, hook, title, thumbnail concept, and trend evidence.