CONTENT CREATION

Flag stale-looking screenshots in docs with a vision model

Scans knowledge-base article screenshots with a Hugging Face vision model to spot outdated UI elements (old logos, deprecated buttons.

CategoryContent Creation
Enginesim
Difficultyintermediate
Triggerschedule
Steps6
Setup~15 min

How it runs

The automated pipeline, trigger to output.

  • TriggerWeekly schedule starts the staleness scan
  • ActionCollect article screenshots from ReadMeReadMeReadMe
  • ActionClassify each image for outdated UI via vision modelHugging FaceHugging Face
  • LogicKeep only images above the staleness threshold
  • ActionCreate a Linear regeneration ticket per flagLinearLinear
  • OutputPost regeneration-ticket digest to SlackSlack

What it does

Reads the screenshots embedded across your knowledge base and runs each through a Hugging Face vision-language model prompted to detect signs of an outdated interface: deprecated branding, removed buttons, legacy color themes, or old navigation. Anything flagged becomes a tracked regeneration task.

When to use it

Use it when you can't easily map docs to live URLs for a pixel diff, but you can still recognize stale UI by eye. The model does the eyeballing at scale across hundreds of images.

How it works

  1. 1A weekly schedule starts the scan.
  2. 2The ReadMe action collects all article screenshots with their source page context.
  3. 3Each image is sent to a Hugging Face vision model with a staleness rubric.
  4. 4A logic step keeps only images scored above the staleness confidence threshold.
  5. 5For each flagged image, a Linear issue is created describing what looks outdated and which article it lives in.
  6. 6A digest of all new regeneration tickets posts to Slack.

Set it up

What you configure once, before turning it on.

  1. 1
    Connect ReadMeAPI docs, changelog, auth.
  2. 2
    Connect Hugging FaceModels, datasets, spaces — the open-source hub.
  3. 3
    Connect LinearIssues, projects, cycles, triage.
  4. 4
    Connect SlackChannels, DMs, threads, mentions.
  5. 5
    Set each agent's modelWe leave models unset so you pick the tier — fast + cheap, or top-quality.
  6. 6
    Tune it to your dataEdit the prompts, filters, and field mappings so it matches how your team works.
  7. 7
    Test, then turn it onRun once against a sample, confirm the output, then enable the trigger.

Run this workflow in your colony.

14-day trial. No DevOps. No Sales call. Provisioned in under a minute.