Devex Review — Workflow
Overview
How the devex-review skill works, step by step.
Source Workflow
Claude skill workflow.
Step-by-step Workflow
devex-review — Workflow
Auto-generated from
SKILL.md. The skill itself is the source of truth; this is a human-readable map of what it does and how it runs.
What it does & when to use
Live developer experience audit. Uses the browse tool to actually TEST the developer experience: navigates docs, tries the getting started flow, times TTHW, screenshots error messages, evaluates CLI help text. Produces a DX scorecard with evidence. Compares against /plan-devex-review scores if they exist (the boomerang: plan said 3 minutes, reality says 8). Use when asked to "test the DX", "DX audit", "developer experience test", or "try the onboarding". Proactively suggest after shipping a developer-facing feature. (gstack) Voice triggers (speech-to-text aliases): "dx audit", "test the developer experience", "try the onboarding", "developer experience test".
Triggers
test the DXDX auditdeveloper experience testtry the onboardingdx audittest the developer experience
How it works
Workflow outline (sections of SKILL.md):
- Preamble (run first)
- Plan Mode Safe Operations
- Skill Invocation During Plan Mode
- Skill routing
- AskUserQuestion Format
- Artifacts Sync (skill start)
- Model-Specific Behavioral Patch (claude)
- Voice
- Context Recovery
- Writing Style (skip entirely if EXPLAIN_LEVEL: terse appears in the preamble echo OR the user's current message explicitly requests terse / no-explanations output)
- Completeness Principle — Boil the Lake
- Confusion Protocol
- Continuous Checkpoint Mode
- Context Health (soft directive)
- Question Tuning (skip entirely if QUESTION_TUNING: false)
- Repo Ownership — See Something, Say Something
- Search Before Building
- Completion Status Protocol
- Operational Self-Improvement
- Telemetry (run last)
- Plan Status Footer
- Step 0: Detect platform and base branch
- SETUP (run this check BEFORE any browse command)
- DX First Principles
- The Seven DX Characteristics
- Cognitive Patterns — How Great DX Leaders Think
- DX Scoring Rubric (0-10 calibration)
- TTHW Benchmarks (Time to Hello World)
- Hall of Fame Reference
- Scope Declaration
- Step 0: Target Discovery
- Step 1: Getting Started Audit
- Step 2: API/CLI/SDK Ergonomics Audit
- Step 3: Error Message Audit
- Step 4: Documentation Audit
- Step 5: Upgrade Path Audit
- Step 6: Developer Environment Audit
- Step 7: Community & Ecosystem Audit
- Step 8: DX Measurement Audit
- DX Scorecard with Evidence
- Boomerang Comparison
- Review Log
- Review Readiness Dashboard
- Plan File Review Report
- GSTACK REVIEW REPORT
- Capture Learnings
- Next Steps
- Formatting Rules
Components
Single-file skill — all instructions live in SKILL.md.
Invoke
- Slash:
/devex-review - Or a natural-language request matching the triggers above.
Source: SKILL.md in this directory.
Execution Logic
The skill executes when its trigger fires (slash command, natural-language match, or direct invocation). It reads its references, applies its rules, and produces the documented outputs.
Edge Cases
See the source skill's references/ and scripts/ folders for edge-case handling.
Failure Handling
A skill failure surfaces as a tool error or a partial output; never a silent skip. Re-run with --verbose (where applicable) for diagnostics.
Integration Notes
- Claude — invoked via the
Skilltool withskill: "devex-review". - Codex — referenced from
AGENTS.mdif mirrored. - Antigravity — referenced from the workspace agent rules if mirrored.
- HQ Project — listed on the landing page Skills section + post-login sidebar.
- MD Project (md.sgnk.ai) — file rendered from
Skills/Devex Review/workflow.md. - Obsidian — file rendered with frontmatter + tags.
Usage Examples
Invoke via slash command or natural language matching the skill description.
Source: ~/.claude/skills/devex-review/WORKFLOW.md