Explain an issue before the fix ships
Describe the symptom, expected behavior and audience. Leave private identifiers out. Use the video to discuss the behavior. Wait for release confirmation before using it in product training.
Mention @ngram-ai on a pull request, issue or release and get an explainer video back in the thread. Ask for the length, shape, style, voice or language in the same comment.
Describe the symptom, expected behavior and audience. Leave private identifiers out. Use the video to discuss the behavior. Wait for release confirmation before using it in product training.
The person who installs the app links an ngram account. People with write access to the repository can then request videos.
Open the ngram-ai listing on GitHub Marketplace, install it on your account or organization and choose the repositories ngram may work in. An organization owner may need to approve the installation.
GitHub sends you to ngram. Sign in, choose Continue with GitHub and confirm your GitHub identity. The link page shows which ngram account pays for the videos. ngram gets no access to your code from this step.
In a new PR or issue comment, write a request such as “@ngram-ai 30 second video for our support team, in Spanish”. For a release, put the mention in its notes before you publish.
ngram replies within seconds, then updates the same reply with the video, usually several minutes later. On a release, the video is added to the notes and attached as a file. Use Open in ngram to edit to keep working on it.
ngram reads the PR description, changed product code and relevant discussion to explain the change. The request can focus on what users will notice.
Ask for a video that explains the problem or proposal in an issue. ngram uses the issue text and conversation as its source.
Include @ngram-ai in the notes before publishing a release. The finished video is added to the notes and attached to the release.
Ask for anything from 5 to 120 seconds; videos default to 30. Add “vertical” for 9:16 or “square” for 1:1, otherwise you get 16:9. Name the audience in your comment.
Name a style such as blueprint, or ask for a calm voice or narration in Spanish. Otherwise ngram designs the look and picks a voice that suits the thread. Your default Brand Kit is always applied.
Only people with write access can trigger a video. Generation uses the credits of the ngram account linked to that installation.
Put the request where the evidence lives. Use a pull request for a changed behavior, an issue for a reported problem and a release for a published set of changes.
In a new PR comment, ask “@ngram-ai explain the new export options to our support team in 20 seconds.” Review the reply against the changed behavior. For a refactor with no visible user impact, keep the explanation in the PR.
Explore product updatesIn a new issue comment, ask “@ngram-ai explain the reported upload failure to triage in 15 seconds.” The issue supplies the description and discussion. If the cause is still unknown, describe the symptom and avoid presenting a proposed fix as shipped.
Explore explainer videosBefore publishing, add “@ngram-ai explain this release to customers, focus on what they can do now” to the release notes. Publishing starts the request. The finished video appears in the notes and as a release asset. Use written notes alone for a list of version bumps.
Explore changelog videosUse the reply’s Open in ngram to edit link to work on the video in the linked account. These features are available in ngram.
Revise the PR explainer’s scenes or pacing in ngram before sharing it with customers.
Explore video editingMake the release explanation easier to follow when a teammate watches without sound.
Explore captionsAsk for a voice in the mention, or change the narration in ngram when a product name needs a different delivery.
Explore voiceoverSet a default Brand Kit in ngram and every @ngram-ai video uses your logo, colors and fonts.
Explore brand kitsDirect attention to the part of a product change that matters most in the edited video.
Explore motion graphicsDevelop a short issue summary into a fuller explanation with more context for the viewer.
Explore explainer creationStart a project in ngram when your source material lives in a changelog or product document.
Create a release walkthrough from written notes when the update is not being published as a GitHub release.
Start from release notesBring several shipped updates together in a changelog video for customers who need the bigger picture.
Start from a changelogTurn the documentation behind a change into a separate walkthrough that explains how to use it.
Start from product docsUse these tools in ngram when your GitHub explanation needs another edit or supporting visual.
Repository contributors can request videos; teammates can use the explanation to understand the change.
Explain a GitHub change in language developers outside the repository can follow.
Ask for the user-facing outcome of a pull request before writing an internal update.
Work with a repository contributor to turn shipped changes into a launch explanation.
Use a short GitHub explainer to show what the team has built to stakeholders.
Create a video through your AI assistant when the technical explanation starts in an assistant conversation.
Explore the MCP serverStart from the documentation page you are browsing when the source is outside a GitHub thread.
Explore the Chrome extensionMention @ngram_ai for an explainer reply in an X conversation, using the bot built for that platform.
Explore the X/Twitter botUse a Zapier workflow when a connected app event should start the video request instead of a GitHub mention.
Explore ZapierPick the ngram entry point that matches where you keep the source material.
GitHub App Request in a repository | ngram workspace Start with your own brief | Chrome extension Start from a webpage | MCP server Ask your AI assistant | |
|---|---|---|---|---|
| Source | A PR, issue or release | A brief and source material | The page open in Chrome | Context supplied by the assistant |
| First action | Mention @ngram-ai | Create a project in ngram | Open the ngram extension | Ask the connected assistant |
| Setup | Install the app and link an account | Sign in to ngram | Install the extension and sign in | Connect the assistant to ngram |
| Result | Video reply or release-note update | A project to review in ngram | A project to continue in ngram | Video result returned to the assistant |
Questions
Install the app from GitHub Marketplace, link your ngram account and mention @ngram-ai in a new comment.