chromaui/chromatic-skills/chromatic-workflow-debug

Claude Code and agent skill by chromaui/chromatic-skills

Trust tier 3
  • community
Source check pending

Checking that the SKILL.md file is still in the repository.

Diagnose Chromatic workflow issues involving git history, baselines, pull request event shape, merge queues, merge-base failures, replacement builds, and patch builds. Use when builds compare against the wrong baseline, an ancestor build is not found, pull request comparisons look wrong, or CI git context does not match Chromatic's expectations.

Install

npx vskill@latest install chromaui/chromatic-skills/chromatic-workflow-debug

SKILL.md

Workflow Debug

Public-safe skill for debugging how Chromatic maps CI, git history, and prior builds into comparisons and baselines.

This package is the single source of truth for:

  • workflow diagnosis codes
  • the evidence ladder
  • the command catalog
  • the diagnosis card output contract
  • documentation topic mapping
  • copy-paste intake prompts
  • public-safe examples and evaluations

Quick start

  1. Read reference/workflow-playbook.md and identify the current workflow phase.
  2. Read reference/diagnosis-taxonomy.md and choose the strongest diagnosis code the evidence supports.
  3. Read reference/evidence-ladder.md and decide whether you can answer now or need one more artifact.
  4. Use reference/command-catalog.md only when the current evidence is insufficient.
  5. Render the result with reference/output-contract.md.
  6. If the user wants follow-up reading, use the topic map in reference/docs-map.md.

Required workflow

1) Start with current evidence

Prefer evidence already provided:

  • the exact Chromatic command or GitHub Action step
  • the relevant warning or error text
  • CI event type and branch information
  • git checkout settings such as shallow history or detached HEAD
  • build URL and the expected comparison target
  • any manual overrides like branchName, repositorySlug, patchBuild, or ignoreLastBuildOnBranch

For repeated approvals, first identify UI Tests versus UI Review. For overlapping builds, collect the upstream acceptance time and first affected intermediate build. Follow the timing checks in reference/workflow-playbook.md before proposing a reset or rerun.

Do not ask for a broad dump if the current evidence already points to one branch of the decision tree.

2) Classify before proposing changes

Choose one primary workflow diagnosis code from reference/diagnosis-taxonomy.md.

If multiple signals exist, prioritize:

  • incorrect or missing git history over downstream symptoms
  • wrong CI branch or PR event shape over later baseline confusion
  • merge-base or patch-build failures over generic comparison complaints
  • replacement-build and rebase behavior over speculative Storybook causes
3) Ask for the smallest next artifact

Ask for exactly one next artifact or one targeted command result.

Typical next artifacts:

  • the exact actions/checkout step or equivalent CI checkout snippet
  • the exact Chromatic invocation line or action inputs
  • git branch --show-current
  • git rev-parse HEAD
  • git log --oneline --decorate -n 10
  • git merge-base <head> <base>
  • the specific Chromatic warning mentioning baselines, merge base, or no ancestor build

Do not jump straight to Storybook flags or TurboSnap suppression flags unless the workflow evidence already cleared the git branch.

4) Stay truthful about the current run

If the current run cannot compute the intended comparison safely, say so plainly.

Examples:

  • a shallow checkout means the current run cannot reason about baselines correctly
  • a pull_request merge commit can point Chromatic at the wrong SHA or branch unless the action passes the correct values
  • a rebased branch may need ignoreLastBuildOnBranch or a fresh baseline path
  • a missing merge base means a patch build path is needed, not a baseline guess
5) Output a diagnosis card every time

Always return the diagnosis card from reference/output-contract.md.

If confidence is still low:

  • name the most likely diagnosis
  • say why confidence is limited
  • request one exact next artifact or command
  • avoid speculative repo surgery

Boundaries

  • Keep the skill public-safe and customer-safe.
  • Ask for one artifact at a time.
  • Do not rely on live external documentation inside the installed skill.
  • Do not recommend unrelated monorepo flags as a first response to git or baseline problems.
  • Do not assume the customer is using GitHub Actions unless the evidence says so, but prefer GitHub Actions examples when you need a concrete pattern.

References and examples

  • reference/workflow-playbook.md
  • reference/diagnosis-taxonomy.md
  • reference/evidence-ladder.md
  • reference/command-catalog.md
  • reference/output-contract.md
  • reference/docs-map.md
  • template.md
  • examples/pr-merge-commit.md
  • examples/rebase-replacement-build.md
  • evaluations/README.md

Security

  • Security scanPassed, 90 of 100Pattern checks on the skill's files.
  • Model reviewPassed

Quality evaluation

  • Task evaluationNot runNo evaluation yet. Run one in Skill Studio.

Source

Popularity

7-day trend score
0

Not refreshed yet.

Works with

Any agent vskill installs to: 56 targets in all. These 39 have a page on this site.

  • Claude Code
  • Cursor
  • GitHub Copilot
  • Windsurf
  • Codex
  • Gemini CLI
  • Cline
  • Amp
  • Roo Code
  • Goose
  • Continue
  • Kilo Code
  • Junie
  • Trae
  • OpenHands
  • Qwen Code
  • AI Agent
  • Tabnine Chat
  • Sourcegraph Cody
  • Amazon Q
  • Firebase Genkit
  • Sweep
  • Devin
  • Phind
  • Mentat
  • Claude Desktop
  • Aider
  • GPT Pilot
  • Plandex
  • bolt.new
  • v0
  • Replit Agent
  • Anse
  • TypingMind
  • BetterChatGPT
  • LibreChat
  • Warp AI
  • Pieces
  • Fig AI
Embed badge
vskill VERIFIED badge
[![vskill VERIFIED](https://verified-skill.com/api/v1/skills/chromaui/chromatic-skills/chromatic-workflow-debug/badge)](https://verified-skill.com/skills/chromaui/chromatic-skills/chromatic-workflow-debug)

Something wrong with this skill or its page? Tell us.