Wdyt collaborate
Participate in a WDYT review by commenting, replying, drawing, signaling readiness, waiting for human activity, or collecting feedback on a running application without adding status noise.From its SKILL.md
npx -y skills add wdyt-page/skills --skill wdyt-collaborateAssembled from the repository path, not quoted from the project. Check it against their README if it does not work.
2 things to look at
- 24 days oldThe repository was created 24 days ago. New is not bad, but a brand new repository carrying a familiar-sounding name is the shape a typosquat arrives in, and there has been no time for anyone else to find a problem with it.
- 0 stars0 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.
SKILL.md
2.1 KB, 437 tokens by cl100k_base, as published. Nobody here has run it
Collaborate in WDYT
Prefer authenticated MCP for owned, workspace, or private pages. Read
../wdyt-access/references/mcp-tools.md for exact tool inputs.
Page collaboration
- Add a comment only to acknowledge completed work, ask a precise question, or anchor feedback to a meaningful location.
- Reply in the existing thread when context exists; do not create parallel status threads.
- Use drawings only when a visual mark communicates more clearly than text.
- Call
wdyt_mark_readyafter a version is actually ready for review. - Use
wdyt_wait_for_reviewin bounded intervals. A timeout means no new activity, not failure. - Delete another participant's contribution only when the human explicitly requests it and the authenticated page role permits it.
Running application review
Use live review when the running application itself is the artifact. Do not rebuild it as static HTML merely to collect feedback.
-
Confirm the target URL loads, uses a stable local port, and permits iframe rendering.
-
Never expose a private development service to the public internet solely for WDYT.
-
Run the CLI doctor, then create the review:
node ../wdyt-cli/scripts/wdyt.mjs live "http://localhost:3000" \ --project "Checkout redesign" --author "Maya's Agent" --json -
Keep the development server alive while the human reviews.
-
Wait only when explicitly asked:
node ../wdyt-cli/scripts/wdyt.mjs wait "<review-url>" --timeout 1800 --json -
Apply feedback to the actual codebase, test it, and record a checkpoint:
node ../wdyt-cli/scripts/wdyt.mjs checkpoint "<review-url>" \ --url "http://localhost:3000" --label "Feedback applied" --json
Read references/live-review.md for target and lifecycle limits. Every outcome should retain the stable review URL and IDs needed for the next action.
What ships with it: 2 files
1.5 KB alongside SKILL.md
agents/
- openai.yaml210 B
references/
- live-review.md1.3 KB