{"id":"19747676-2ed3-44b2-b8b1-830aa45b88c0","slug":"ai-build-kit","kind":"skill","name":"ai-build-kit","summary":"A workflow for building and running internal tools with a coding agent.","owner":"gwpicard","repoUrl":"https://github.com/gwpicard/ai-build-kit","license":"MIT","licenseTier":"permissive","trust":"community","stars":2,"match":null,"fewStars":true,"format":"skill","createdAt":"2026-09-11T08:48:51.000Z","addedAt":"2026-10-21T11:34:06.727Z","logoUrl":"https://avatars.githubusercontent.com/u/7591224?u=220ddffeff6ed9b0d827c90921f074dd6f0b24ae&v=4","linkOnly":false,"collection":null,"skillPath":null,"categories":["databases","developer-tools"],"tags":["ai-assisted-development","claude-code","codex","coding-agents","internal-tools","nextjs","supabase","vibe-coding"],"runtime":"binary","local":true,"cloudCapable":false,"cloudCapability":"unknown","cloudReadyOnPlatform":false,"needsApiKey":false,"reviewed":false,"syncedAt":"2026-10-26T18:46:42.017Z","sample":false,"url":"https://conduits.dev/tools/ai-build-kit","language":"Shell","transports":["stdio"],"install":{},"requirements":{"env":[],"upstreamMayCharge":false},"tools":null,"readme":"# AI Build Kit\n\n[![Licence: MIT](https://img.shields.io/github/license/gwpicard/ai-build-kit)](LICENSE)\n[![Latest release](https://img.shields.io/github/v/release/gwpicard/ai-build-kit)](https://github.com/gwpicard/ai-build-kit/releases/latest)\n\n**[Site](https://gwpicard.github.io/ai-build-kit/)** · [Releases](https://github.com/gwpicard/ai-build-kit/releases)\n\nAI Build Kit helps people who do not write code build and run tools with a\ncoding agent. It suits team leads, operations staff, analysts, product\nmanagers and founders who are comfortable with technology. Developers new to\nagent-led work can use it too.\n\nThe six commands are:\n\n- `/setup-ai-build-kit` prepares a project.\n- `/shape` turns an idea or fault into a ready piece of work.\n- `/implement` builds a ready piece and asks before merging by default.\n- `/setup-hosting` sets up or checks how the tool runs live.\n- `/maintain` checks the records and live copy, and runs upkeep when due.\n- `/what-now` says where the project stands and what to do next.\n\nAutomatic merging is off by default. The masterplan records the choice and its\nlimits under `Runs on its own:`.\n\nThe kit does not host tools or store project records. It is for internal tools\nwith a manual fallback. Passing checks do not prove that software is safe or\nright for your purpose. The kit cannot replace a professional developer or\nprovide legal, medical, financial or security advice. You use it at your own\nrisk. The terms of the services you run a tool on, such as a host or a database\nplan, are between you and those services, and you are responsible for reading\nand following them.\n\n## Quick start\n\nChoose one route for each project. For the Claude Code terminal app alone,\nrun these commands from your project folder:\n\n```bash\nclaude plugin marketplace add gwpicard/ai-build-kit\nclaude plugin install ai-build-kit@ai-build-kit --scope local\n```\n\nStart Claude Code there and type `/ai-build-kit:setup-ai-build-kit`.\n\nFor another agent, or a project using several agents, run:\n\n```bash\nnpx skills add gwpicard/ai-build-kit\n```\n\nChoose the agents you use and install all the kit's skills. Then ask your\nagent: \"Run the setup-ai-build-kit skill.\"\n\nFor Codex on a Mac, start each session from your project folder with\n`python3 .agents/skills/setup-ai-build-kit/scripts/codex-with-github.py`.\nSetup explains any extra tools your recipe needs. [Manual installation and\nAgent Plugins](docs/COMPATIBILITY.md#choose-one-installation-route) are also\navailable.\n\n[Read the AI Build Kit site](https://gwpicard.github.io/ai-build-kit/) for the\nworkflow, the rules and checks, supported agents, design, releases and sources.\n\n## A sample session\n\nThis shortened example comes from a replayed session on an office noticeboard\napp, using Claude Code. The tool is already live and a change is ready to merge.\n\n> **Person:** `/implement` I opened the preview and the sign-in button says\n> \"Email me a sign-in link\", which is right. Merge it and put it live.\n>\n> **Agent:** All four tests and the project check pass. I've added a dated\n> changelog line. The review has no unresolved concern, so I'm asking Claude\n> Code to merge it. Claude Code will show you a confirmation box before the\n> merge runs.\n\nAfter the merge, the agent reads the live page:\n\n> **Agent:** The sign-in button at the live address now says \"Email me a\n> sign-in link\". The live copy runs this change. This read only checks the button text.\n>\n> **Person:** Priya still sees the old \"Continue\" button. Can you put it out again?\n>\n> **Agent:** The live page already shows the new button. Try a full refresh\n> in Priya's browser. I have not deployed it again.\n\n## Contributing and licence\n\nRead [CONTRIBUTING.md](CONTRIBUTING.md) to report a problem or suggest a\nchange. The kit is provided as is, without warranty, under the [MIT licence](LICENSE).\n"}