How Church Media Kit turns a public YouTube sermon into a Sermon Kit

You have a public sermon link and a list of posts your church could use this week. It would be easy to paste the link, accept whatever appears, and call the job finished.

Church Media Kit does not work that way. The reviewed build moves through checkpoints. It checks the source, asks your team what to make, and prepares draft designs. Then a person has the last word.

Some checks are automatic. Pastoral, factual, legal, and publishing decisions still belong to the church. This tour follows the full path, including the places where the work should stop.

Start with a supported public YouTube sermon

The intake begins with a public YouTube watch URL. The system validates the address and rebuilds a canonical YouTube link before it uses the source.

Next comes a rights confirmation. The person submitting the sermon states that the video was produced by or licensed to the church and that they are authorized to repurpose it.

That statement matters, but it is still a statement. Church Media Kit does not investigate ownership, grant a licence, or obtain permission from YouTube. Your church must know whether it may reuse the sermon, speaker image, music, scripture translation, testimony, photography, and any other material that could appear in a post.

Stop here if the person submitting the source cannot make that decision responsibly.

Read the caption result before choosing content

Church Media Kit checks the video for captions after the rights confirmation. The result can take one of several forms.

• Captions are available.
• No captions were found.
• The source is invalid.
• The caption check is temporarily unavailable.

An invalid source cannot become a Sermon Kit. A temporary failure calls for a retry. It is not evidence that captions are missing.

Available captions give the workflow sermon text to work with. They do not prove that every word is right. Proper names, scripture references, uncommon theological terms, accents, overlapping speech, and poor audio can all cause errors.

YouTube says a transcript is available only when a video has captions. Viewers can jump from a transcript line to the matching moment in the video. That makes YouTube's transcript view useful for spot checks.

If no captions are found, the YouTube Sermon Content intake does not transcribe the sermon audio. The current intake can instead use a manually supplied transcript, manuscript, or notes with at least 40 words. Check that text against the recording before relying on it. If no suitable text is available, add or repair the video’s captions in YouTube, or use another verified source. YouTube documents caption file, manual-entry, and automatic-caption options.

Choose direction, length, and the actual content set

Once you have usable source text, the setup asks how the drafts should sound. The reviewed build offers four directions: Warm and Pastoral, Bold and Convicting, Reflective and Quiet, and Uplifting and Joyful. Current marketing fixes caption length to balanced.

These controls guide the draft. They do not teach the system your church's voice. A warm selection cannot know how your pastor speaks to a grieving family. A bold selection cannot decide whether a sharp line preserves the sermon's invitation or strips away its care.

The next choice is the content set. You can select counts for Quote, Big Idea, Teaching, Announcement, Prayer, Question, Devotion, Testimony, Sermon Recap, and Stat Impact posts.

Choose what the source can support, not the largest possible set. A sermon with no event details should not be forced into an announcement. A private story should not become a testimony card because the format is available. A number should not become a Stat Impact post unless the sermon states it and a reviewer can verify what it means.

The useful question is not "How many posts can we get?" It is "Which posts help our people understand or act on this message without changing it?"

Generation becomes a saved job

Starting generation creates a saved job with its own identity. In the reviewed workflow, that job can be queued, running, succeeded, failed, needs-attention, or in reconciliation. The workflow also tracks the selected content sections separately.

That design matters when work is interrupted. A browser refresh does not need to pretend that nothing happened. A section that reaches a ready checkpoint can remain distinct from a section that failed. Some failed sections can be selected for another attempt. An uncertain provider result can move to reconciliation instead of triggering a blind repeat.

Those are recovery controls, not a recovery guarantee. They do not prove that every attempt finishes, that a provider call runs exactly once, or that generation will take a particular amount of time. A needs-attention state means a person needs to decide what to retry or abandon. A reconciliation state means the system cannot safely infer the outcome and will not resolve the uncertainty by inventing one.

The church's decision is simple here. Do not start a second unrelated run just because the first one looks slow. Check the saved job state first. If the interface asks for manual review, treat that as a real stop.

The system builds evidence, then applies narrow checks

When captions are available, the workflow turns caption segments into bounded excerpts with timestamps. It uses those excerpts to prepare a sermon brief and a plan for each selected content section.

The reviewed build applies several automatic checks. Among other things, it requires a section plan to use evidence from the sermon brief, rejects generated evidence that falls outside the section plan, checks required design fields, enforces copy limits for design slots, and verifies that a preview has the expected pages and text roles. The generation instructions also distinguish exact quotes from paraphrases and tell the system not to invent names, event details, addresses, or scripture references.

These checks catch specific structural problems. They do not prove that a post is true, wise, kind, accessible, or faithful to the sermon's purpose.

An exact line can still be misleading when removed from its surrounding argument. A paraphrase can pass a character limit while making the pastor sound more certain than the sermon was. A scripture reference can appear in the captions and still be misheard. An announcement can be complete in shape and out of date in fact.

The source evidence helps a reviewer find what to check. It does not replace the check.

Review the generated Sermon Kit

After the selected sections are ready, the Sermon Kit groups the generated static designs by content type. A design may have alternatives. The overview can mark a design as needing review and can block download when the selected pages are not ready.

"Ready" has a narrow meaning. It means the product's current structural checks have not found a blocking issue. It does not mean that a pastor approved the theology, a communications lead approved the voice, a ministry owner confirmed the facts, or an accessibility reviewer approved the page.

Review each design against the source before choosing an alternative. Use this order:

1. Confirm that every quotation is word-for-word. Remove quotation marks from a paraphrase.
2. Read enough of the surrounding transcript to recover the sermon's actual point.
3. Verify names, dates, locations, service times, statistics, scripture references, and calls to action from a current source.
4. Remove prayer requests, safeguarding details, testimonies, or personal stories that do not have the right consent for a social post.
5. Check whether the image has the right subject and permission.
6. Read the page at phone size. Check type size, contrast, and whether the important words survive the design.
7. Write or correct the image alternative so it describes what the image contributes.

The W3C advises authors to give informative images a short alternative that conveys their essential information, while decorative images use an empty alternative. Its images tutorial is a practical reference. WCAG 2.2 sets a 4.5:1 minimum contrast ratio for ordinary text and 3:1 for large text, with defined exceptions.

Edit the static pages in Studio

Studio is where the team corrects the draft. The reviewed editor supports changes to text, images, shapes, backgrounds, page order, and layers, with undo and redo. The point is not to decorate every page. It is to make the smallest changes that bring the post back to the source, the church's voice, and the intended reader.

The editor does not automatically apply a saved church Brand Kit. Branding remains a manual design decision in this reviewed build. It is a static-design editor and does not claim parity with a general-purpose design platform.

Delete weak work. A Sermon Kit is not better because every generated idea survives. If a design needs too much explanation, uses a thin source fragment, repeats another post, or creates a pastoral problem, removing it is often the best edit.

Export only the pages the church approves

The export flow lets the user choose all pages, the current page, individual pages, or a page range. A selected page that still needs review blocks the export action. One approved page can be downloaded as a static image. Several approved pages can be collected in a ZIP.

Export is not approval, and it is not publication. It creates files that your church can move into its own approved publishing process. This tour makes no promise about live social destinations, direct carousel publishing, or automatic posting.

Keep the approved copy, image permissions, alt text, destination, and publication time together. If a fact changes after export, update the design before anyone posts it.

Compare the job, not the feature count

Other sermon-repurposing products advertise broader workflows. Sermon Shots claims transcription, video clips, images, templates, and direct social sharing. ChurchSocial Sermon Studio claims clips, written content, scheduling, and team collaboration. Pulpit AI claims audio, video, and document uploads with a larger set of ministry resources.

Those are vendor claims, not independent tests. They expose different buying decisions. If your church needs captionless audio transcription or a particular publishing workflow, compare a product that can demonstrate that job now. External publishing is not included in the current Church Media Kit subscription.

Church Media Kit's reviewed path is narrower. It starts with a public YouTube sermon, depends on source evidence, prepares static drafts, and keeps a human review between generation and export.

Use one ordinary sermon to verify the product

A polished sample tells you very little. Bring an ordinary authorized sermon to a live demonstration. Choose one with a proper name, a scripture reference, a paraphrase, an announcement, and at least one section that should not become a post.

Ask to see the whole path:

• Submit the real public URL and show the rights confirmation.
• Open the captions and compare one line with the video.
• Use a captionless source to show the manual-text fallback: supply a verified transcript, manuscript, or notes of at least 40 words and compare it with the recording. If no verified text is available, add or repair captions before retrying.
• Select a small content mix instead of every format.
• Follow the saved job through an interruption or needs-attention state.
• Trace one exact quote and one paraphrase back to their evidence.
• Correct a factual detail and replace one unsuitable image.
• Leave one page unresolved and confirm that export stays blocked.
• Approve a different page and export only that page.

The test should feel slightly inconvenient. That is where the truth appears. A product tour is useful when it shows the failure state, the correction, and the final handoff, not only the clean result.

The fit in one sentence

Church Media Kit may fit a church that has an authorized public YouTube sermon or verified text source, wants editable social drafts, and has people available to review the source and approve each page. See the current Sermon to social feature and pricing.

It does not fit when the YouTube Sermon Content intake must transcribe audio and captions or verified supplied text are unavailable, or when the church needs automatic branding or approval, external publishing from the subscription, or guaranteed outcomes.

The product can prepare material and expose checkpoints. Your church still decides what the sermon means, what the public should see, and whether a page is ready to carry the church's name.