Agent-first development toolkit. 5 skills, 8 commands, 4 hooks, 1 MCP server. Bootstrap knowledge bases, shape requirements, breadboard systems, browser automation, and enforce standards.
Find and fix design smells in breadboards. Use after /breadboarding to validate affordance boundaries, naming, and wiring correctness.
Map a system into Places, UI affordances, Code affordances, and wiring. Use to map existing systems or design new ones from shaped parts.
Add optional ESLint-based mechanical enforcement to steer agents (JS/TS repos).
Bootstrap an agent-first knowledge base in this repository
Onboard an existing repository into the harness-engineering knowledge base and fill in gaps using an automated audit.
This skill manages Git worktrees for isolated parallel development. It handles creating, listing, switching, and cleaning up worktrees with automatic .env file copying and shared-context configuration. Use when spinning up parallel feature branches, code review in isolation, or multi-agent development.
Bootstrap an agent-first knowledge base in any repository. This skill should be used when initializing a new project or adding structured knowledge scaffolding to an existing repo. Supports configurable base directory, new/existing repos, Claude settings with plan directory, and bundled agent capabilities. Invoke via /harness-init.
This skill manages Git worktrees for isolated parallel development. It handles creating, listing, switching, and cleaning up worktrees with automatic .env file copying and shared-context configuration. Use when spinning up parallel feature branches, code review in isolation, or multi-agent development.
Automates browser interactions for web testing, form filling, screenshots, and data extraction. Use when the user needs to navigate websites, interact with web pages, fill forms, take screenshots, test web applications, or extract information from web pages.
Shape Up methodology adapted for LLMs. Three phases: /shaping (iterate on requirements and solution shapes with fit checks), /breadboarding (map systems into Places, UI affordances, Code affordances, and wiring), /breadboard-reflection (find and fix design smells in breadboards). Use when defining problems, exploring solutions, mapping systems, slicing into vertical scopes, or validating breadboard designs.
Modifies files
Hook triggers on file write and edit operations
External network access
Connects to servers outside your machine
Own this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimOwn this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimBased on adoption, maintenance, documentation, and repository signals. Not a security audit or endorsement.
harness-engineering
Agent-first development toolkit for Claude Code
Install · Skills · Commands · Hooks · Contributing
Bootstrap knowledge bases, shape requirements, breadboard systems, and enforce standards — all from inside Claude Code.
Built on the methodology behind ~1M LOC across 1,500 PRs with zero manually-written code.
| Component | Count | Summary |
|---|---|---|
| Skills | 5 | harness-engineering, git-worktree, shaping, tdd, playwright-cli |
| Commands | 8 | /harness-init, /harness-shape, /shaping, /breadboarding, and more |
| Hooks | 4 | Lint-on-save, shaping ripple checks, plan capture, plan pipeline |
| MCP Servers | 1 | context7 — framework documentation lookup |
Inside Claude Code:
/plugin marketplace add lattice-technologies-inc/harness-engineering
/plugin install harness-engineering@harness-engineering-marketplace
git clone https://github.com/lattice-technologies-inc/harness-engineering.git
cd harness-engineering
# Load for a single session
claude --plugin-dir .
/plugin marketplace add /path/to/harness-engineering
/plugin install harness-engineering@harness-engineering-marketplace
Scope options: Add
--scope projectto share with your team (writes to.claude/settings.json) or--scope localfor this repo only (gitignored).
/harness-init # Bootstrap knowledge base in your repo
/harness-shape # Fill placeholders with real project context
/shaping # Shape a feature (requirements + solution options)
/breadboarding # Map system into affordance tables
New repo:
/harness-init --> /harness-shape --> /harness-eslint (optional)
Existing repo:
/harness-init --> /harness-onboard --> /harness-shape --> /harness-eslint (optional)
Bootstrap an agent-first knowledge base in any repository. Creates AGENTS.md, structured docs, golden principles, quality scoring, and mechanical enforcement.
What gets scaffolded:
AGENTS.md — table of contents (not an encyclopedia)ARCHITECTURE.md — domain mapdocs/ — golden principles, quality scores, tech debt tracker, design docs, plans pipeline.claude/ — commands, hooks, settingsscripts/harness/knowledge-check.sh — CI-ready structural checksShape Up methodology adapted for LLMs. Three commands covering the full pre-code workflow:
| Command | Phase | Output |
|---|---|---|
/shaping | Define problem + solution options | Requirements (R), Shapes (S), fit checks |
/breadboarding | Map into concrete affordances | Places, UI/Code affordances, wiring, vertical slices |
/breadboard-reflection | Validate design | Smell detection, naming tests, wiring coherence |
Test-driven development with vertical red-green-refactor loops. Enforces behavior-over-implementation testing, boundary-only mocking, and incremental tracer bullet workflow.
Manage Git worktrees for isolated parallel development. Handles env file copying, .worktree-config for multi-agent coordination, and cleanup.
Playwright-powered browser automation CLI. Full browser control with snapshots, tabs, cookies, network mocking, tracing, and CDP support (connect to Arc, Chrome, Edge).
| Command | Description |
|---|---|
/harness-init | Bootstrap knowledge base scaffolding |
/harness-shape | Fill placeholders with real project context |
/harness-standards | Alias for /harness-shape |
/harness-onboard | Audit existing repo and generate gap-fill plan |
/harness-eslint | Add agent-readable ESLint enforcement (JS/TS) |
/shaping | Iterate on requirements and solution shapes with fit checks |
/breadboarding | Map systems into affordance tables with wiring |
/breadboard-reflection | Validate breadboard design for smells |
| Hook | Trigger | What it does |
|---|---|---|
| lint-on-save | PostToolUse on Edit/Write | Runs ESLint, ruff, or shellcheck and feeds errors back to the agent |
| shaping-ripple | PostToolUse on Edit/Write | Checks if shaping doc edits need to ripple to related docs |
| plan-capture | PostToolUse on ExitPlanMode | Reminds agent to file the approved plan in docs/plans/ |
| plan-pipeline | Stop | Verifies plans are in the correct pipeline stage before session ends |
npx claudepluginhub lattice-technologies-inc/harness-engineering --plugin harness-engineeringFocused agentic engineering workflow: design-doc, spec, plan, implement, tdd, refactor, review, address-pr-feedback, browser-verify, explain-visually, compress, branch, and commit.
Personal Claude Code / Codex skills — spec-driven development, agent teams, browser automation, productivity, and setup helpers. Install with `npx skills add devstefancho/claude-plugins`.
Compound Engineering workflow: PRD-driven sprints, isolated worktrees, hook-enforced safety, automated learning. Skills become /vini-workflow:plan, /vini-workflow:compound, etc.
HarnessFlow — From idea to shipped product: high-quality engineering workflows for AI agents. Spec-anchored SDD, gated TDD, evidence-based routing, independent reviews, and formal closeout.
Requirements-driven development workflow with quality gates for practical feature implementation
Harness Engineering framework - skills, agents, and commands for safe, reviewable, incremental agent-driven development. Includes RPEQ workflow (Research, Plan, Execute, QA), ast-grep setup, and codebase analysis tools.