Skip to content

docs: add Agent Canvas visuals from image request #768 - #769

Draft
DevinVinson wants to merge 1 commit into
mainfrom
dv/docs-agent-canvas-images-issue-768
Draft

docs: add Agent Canvas visuals from image request #768#769
DevinVinson wants to merge 1 commit into
mainfrom
dv/docs-agent-canvas-images-issue-768

Conversation

@DevinVinson

Copy link
Copy Markdown
Member

Closes #768

Summary

  • Adds a focused screenshot of the Agent Canvas Add extension dialog.
  • Places it immediately after the opening instruction in canvas-extensions.mdx.
  • Uses synthetic source, ref, and repository-path values to clarify the Git Repository workflow.

Source audit

  • Agent Canvas source SHA: f26d734a848297d8dcf460b0bb739174e76511f0
  • Documentation base SHA: 5a75b32c7f1e93811e8ccf440ad577307cc35bd6
  • Theme: OpenHands-Neutral

Accepted and omitted images

Accepted

  • agent-canvas-add-extension.png — focused Add extension dialog showing the synthetic Source, Ref, and Repo path values with the enabled action. PNG, 560 × 472.

Omitted

Image QA and privacy

  • Captured through the production static frontend, real Agent Server, real automation backend, scripted mock LMM, isolated synthetic state, and Chromium.
  • Chromium capture used a 1440 × 1000 viewport, scale 1, English, and America/New_York.
  • Verified OpenHands-Neutral, suppressed animations/transitions/caret, one visible dialog, all requested synthetic values, and enabled primary action.
  • Verified the focused crop has no browser chrome, secrets, personal data, real local paths, private URLs, conversation identifiers, errors, skeletons, banners, unexpected modals, or toasts.
  • First candidate was rejected because Chromium inherited a noncompliant viewport. The permitted recapture explicitly set and verified the required viewport and passed QA.
  • Inspected the final candidate and one-image contact sheet outside the repositories.

Validation

  • Playwright mock-LLM capture flow — passed (exit code 0).
  • Theme assertion (openhands-neutral) and theme style presence — passed.
  • Dialog count and scoped forbidden-text/privacy scan — passed.
  • PNG decode/format/dimensions — passed (560 × 472, RGB, non-interlaced).
  • Image reference and orphan check — passed (one in-scope reference).
  • Target-page image path/load audit — passed.
  • Expected tracked-change scope — passed (one MDX page and one PNG).
  • git diff --check — passed.
  • mint broken-links — passed (success no broken links found).
  • uv run --with pytest --with requests pytest -q tests/ — failed: 35 passed, 2 setup errors.

Known test failures

  • uv run --with pytest --with requests pytest -q tests/ — 35 tests passed; two pricing-documentation tests errored while fetching https://raw.githubusercontent.com/OpenHands/OpenHands/main/openhands/utils/llm.py, which returns HTTP 404. This appears unrelated to the Agent Canvas MDX/PNG changes because the failure is an existing external source-path lookup in tests/test_pricing_documentation.py.

This pull request was created by an AI agent (OpenHands) on behalf of the user.

Co-authored-by: openhands <openhands@all-hands.dev>
@mintlify

mintlify Bot commented Aug 29, 2026

Copy link
Copy Markdown

Preview deployment for your docs. Learn more about Mintlify Previews.

Project Status Preview Updated (UTC)
all-hands-ai 🟢 Ready View Preview Aug 29, 2026, 12:45 PM

💡 Tip: Enable Workflows to automatically generate PRs for you.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

automation: Agent Canvas documentation image requests — 2026-08-29

1 participant