SkillAgentSearch skills...

rocky

Agentic-assisted engineering for teams and solo devs

Install / Use

claude mcp add hellozheat -- npx -y github:hellozheat/rocky

If the server publishes to npm under a different name, use that package instead — check the repo README.

About this skill
🔌

MCP Server

Model Context Protocol server

Quality Score

70/100

Supported Platforms

Claude Code
Claude Desktop

Tags

Our assessment of rocky

rocky scores 70/100 on our quality scale, 3833rd of 4,658 Development & Engineering skills we index.

Its MCP Server is 6.2 KB long, well organised into 20 sections with 6 code examples: a thorough specification that gives an agent plenty to work with.

It has 3 GitHub stars, so there is little community track record yet; judge it on its content.

Substance
29/30
Structure
20/20
Description
8/15
Adoption
3/20
Freshness
11/15

Maintenance, license and trust

  • The repository was last updated about 5 months ago. That is recent enough to be usable, but agent tooling moves fast, so check the instructions against your agent's current version.
  • No license is declared. By default that means all rights are reserved: you can read it, but reusing or redistributing it is not clearly permitted. Ask the author before building on it commercially.
  • Its trust signals score 73/100, with 3 cautions from licensing, adoption, age or documentation. These come from repository metadata, not a code audit — read the skill file before letting an agent act on it.

rocky compared with similar skills

All 4 of these similar skills score higher than rocky; compare them before choosing.

SkillScoreStarsUpdatedFormat
rocky (this skill)by hellozheat7035mo agoMCP Server
Agent-Reachby Panniantong100106.6k5d agoCLAUDE.md
taste-skillby Leonxlnx100106.0ktodayCLAUDE.md
headroomby headroomlabs-ai10078.3ktodayCLAUDE.md
ai-job-searchby ai-job-search-dev10067.6k3d agoCLAUDE.md

Frequently asked questions

How do I install rocky?
Run claude mcp add hellozheat -- npx -y github:hellozheat/rocky. The install tabs above show the steps for each supported agent.
Which AI agents does rocky work with?
It is written for Claude Code and Claude Desktop, as a MCP Server file. Other agents that read the same format can often use it too.
Is rocky safe to use?
It declares no license and scores 73/100 on trust signals. Skills are instructions an agent will follow, so read the file before installing it and do not approve commands you do not understand.
Is rocky still maintained?
The repository was last updated about 5 months ago. That is recent enough to be usable, but agent tooling moves fast, so check the instructions against your agent's current version.

Rocky — Agentic-assisted engineering for teams and solo devs

Rocky hero

License: MIT MCP Endpoint Inspector

Rocky is an MCP server for AI-assisted engineering workflows. It gives your assistant a shared handbook (agents + rules), one devkit gateway tool, and a pre-PR quality gate so outputs are reviewable before opening a pull request.

Rocky is agentic-assisted, not autonomous: your model still writes code in your repo, and Rocky provides the standards, routing, and checks.

🚀 Quick connect

MCP endpoint: https://userocky.zheat.xyz/mcp
Inspector: https://userocky.zheat.xyz/inspector

Claude Code

claude mcp add --transport http "rocky" https://userocky.zheat.xyz/mcp

Cursor

Add to ~/.cursor/mcp.json:

{
  "mcpServers": {
    "rocky": {
      "url": "https://userocky.zheat.xyz/mcp"
    }
  }
}

VS Code / VS Code Insiders / ChatGPT

  • Open userocky.zheat.xyz/mcp and use one-click client buttons, or
  • Add https://userocky.zheat.xyz/mcp manually in your MCP connector settings.

Why teams use Rocky

  • Pre-PR quality gate: run lint, tests, and review heuristics before PR creation.
  • Token-efficient tooling: one devkit gateway action contract instead of fragmented tools.
  • Codebase discovery first: graphify-first flow, optional Understand Anything graph if present.
  • Consistent implementation style: shared agents and rules exposed as MCP resources.
  • Safer repo operations: path scoping via DEVKIT_ALLOWED_REPO_ROOTS and allowlisted command execution.

Snapshot: features, agents, skills

Features

  • devkit unified gateway + standalone repo/web helper tools.
  • pre_pr_quality_gate with ready / not_ready verdicts.
  • Codebase discovery flow: Understand Anything (if present) -> graphify -> folder walkthrough.
  • Handbook resources exposed over MCP (devkit://handbook/...).
  • Built-in prompts: devkit-start-task, devkit-review-code, devkit-before-pr, devkit-learn-the-stack.

Agents (24 total)

  • Architecture/app: react-hexagonal, nextjs-developer, node-api-developer, nestjs-hexagonal, fastapi-hexagonal, typescript-library-developer.
  • Testing/quality: vitest-writer, playwright-writer, storybook-writer, test-writer-nestjs, test-writer-python, tester-qa, pr-quality-gate, code-reviewer.
  • Discovery/workflow: codebase-discovery, graphify-local-project, graphify-codebase-understanding, understand-anything-onboarding, coverage-and-review-workflow.
  • Support: product-owner, documentation-writer, code-simplifier, tailwind-ui-developer, k3s-devops.

Skills (26 total, compact groups)

  • Build by stack: react-hexagonal, nextjs-app-router, node-api-layered, nestjs-hexagonal, fastapi-hexagonal, typescript-library.
  • Testing/review: vitest-react, playwright-e2e, storybook-react, jest-nestjs, pytest-hexagonal, linter, githubpr, feature-implementation.
  • Ops/security/perf: trivyfix, sonarfix, popeyescan, performance-audit, dbanalyze, async-python-patterns.
  • Product/design: product-requirements, frontend-design, web-design-guidelines, brainstorming, tailwind-ui, vercel-react-best-practices.

Usage

Use natural language in your host app; the model calls devkit actions under the hood.

Example prompt

Use Rocky: read devkit://how-it-works, list_handbook, and the matching stack agent.
Implement changes directly in this repo, then run pre_pr_quality_gate before we finish.
Do not only call MCP tools without editing files.

Typical flow

  1. Discover conventions (list_handbook, stack-specific agent).
  2. Implement in the repo (small, reviewable diffs).
  3. Verify (yarn lint:fix, yarn test, yarn build in target project).
  4. Gate (pre_pr_quality_gate) until verdict is ready.
  5. Open PR (repo_open_pr) if needed.

Integrations and clients

  • MCP hosts: Cursor, Claude Code/Desktop, VS Code, VS Code Insiders, ChatGPT (Developer Mode).
  • Inspector: built-in interactive MCP inspection at /inspector.
  • Team workflow: works with existing repo conventions, CI, and review processes.

Local development

git clone https://github.com/hellozheat/rocky.git
cd rocky
yarn install
yarn dev

Build and verify:

yarn lint:fix
yarn test
yarn build

Deploy

yarn install
yarn build
yarn deploy

Set production env vars from config/env.example:

  • MCP_URL
  • PORT
  • DEVKIT_ALLOWED_REPO_ROOTS
  • GITHUB_TOKEN (only if using repo_open_pr)

Documentation

Core concepts (quick reference)

  • Gateway model: call devkit with an action for handbook, analysis, quality gate, and repo operations.
  • Discovery policy: prefer Understand Anything graph when present, otherwise graphify artifacts, otherwise folder walkthrough.
  • Prompt resources: devkit-start-task, devkit-review-code, devkit-before-pr, devkit-learn-the-stack.

For full action-level references and MCP resource details, use:

  • devkit://capabilities
  • devkit://tool-composability
  • devkit://path-scope-policy

Contributing

Contributions are welcome. Please open an issue or PR with clear scope and reproducible steps.

If your change affects behavior, include:

  • before/after behavior summary,
  • tests or rationale for test coverage,
  • any handbook/rule updates needed for agent consistency.

License

MIT

Related Skills

View on GitHub
GitHub Stars3
CategoryDevelopment
Updated5mo ago
Forks1

Languages

TypeScript

Trust signals

73/100

From repository metadata: license, adoption, age and documentation. Not a code audit — see the Safety scan above for what the skill file itself contains.

1 medium2 low1 info