From harness
Use when the user says "/harness:skillflow", "/harness", "start a skillflow", or needs help choosing between shape, brief, ground, design, prototype, and report skills for this project.
npx claudepluginhub lukaleskovsek/makers-and-breakers-claude-marketplace --plugin harnessThis skill uses the workspace's default tool permissions.
Entry point for the Harness pre-implementation skill bundle. Route users to the right skill for what they need.
Guides Next.js Cache Components and Partial Prerendering (PPR) with cacheComponents enabled. Implements 'use cache', cacheLife(), cacheTag(), revalidateTag(), static/dynamic optimization, and cache debugging.
Migrates code, prompts, and API calls from Claude Sonnet 4.0/4.5 or Opus 4.1 to Opus 4.5, updating model strings on Anthropic, AWS, GCP, Azure platforms.
Automates semantic versioning and release workflow for Claude Code plugins: bumps versions in package.json, marketplace.json, plugin.json; verifies builds; creates git tags, GitHub releases, changelogs.
Entry point for the Harness pre-implementation skill bundle. Route users to the right skill for what they need.
.harness/kb/stack-profile.md. If missing, tell the user to run /harness:init first.harness_status to verify. If it fails, tell the user to start the daemon.| Skill | Command | Best for |
|---|---|---|
| Shape | /harness:shape | "Can this project do X?" — explore feasibility, shape ideas |
| Brief | /harness:brief | "Document this idea" — create structured briefs |
| Ground | /harness:ground | "Spec this out" — grounded specs with codebase reality checks |
| Design | /harness:design | "How should this look?" — UI/UX component design + HTML variants |
| Prototype | /harness:prototype | "Build a demo" — working prototypes following project conventions |
| Report | /harness:report | "Summarize this" — stakeholder-ready documentation |
Based on user intent:
| Signal | Route to |
|---|---|
| "Can we do X?", "Does the system support..." | /harness:shape (quick mode) |
| "I have an idea", "What if we...", "Let's brainstorm" | /harness:shape (guided mode) |
| "Write a brief", "Document this idea" | /harness:brief |
| "Spec this out", "What would it actually take" | /harness:ground |
| "Design the UI", "How should this look" | /harness:design |
| "Build a prototype", "Show me this working" | /harness:prototype |
| "Create a report", "Summarize the status" | /harness:report |
| "I have a brief already" | /harness:ground (with brief path) |
| "I have a spec, build it" | /harness:prototype or hand off to implementation |
Before routing, scan documents/shapings/ (or the project's target_dir) for recent files:
ls -lt documents/shapings/ | head -10
If recent files exist, mention them: "There's a [type] from [date] about [topic]. Continue from there?"
When the user wants to start a full workflow:
harness_create_skillflow with a title and the repo_idharness_open_in_dashboard so the user can see progress in the browser/harness:shape for new ideas, /harness:ground for existing specs)Skills pass context through files in the target directory. Each saved artifact has frontmatter with a next-skill field. Users manually invoke the next skill — no automatic chaining.
Typical flow:
/harness:shape → conversation → suggests /harness:brief/harness:brief → saves brief → suggests /harness:ground or /harness:prototype/harness:ground → saves spec → suggests /harness:design or /harness:prototype/harness:design → saves design doc + HTML variants → suggests /harness:prototype/harness:prototype → creates working demo → suggests /harness:reportThis router is a traffic cop, not a gatekeeper. If the user knows which skill they want, suggest it immediately. Only ask clarifying questions when intent is genuinely ambiguous.