Multi-model AI orchestration workflow for Claude Code
# Add to your Claude Code skills
git clone https://github.com/Fredasterehub/kilnLast scanned: 5/30/2026
{
"issues": [],
"status": "PASSED",
"scannedAt": "2026-05-30T15:43:57.000Z",
"npmAuditRan": true,
"pipAuditRan": true
}kiln is an open-source ai agents skill for AI coding assistants such as Claude Code, Codex CLI, and ChatGPT, built by Fredasterehub. Multi-model AI orchestration workflow for Claude Code. It has 202 GitHub stars.
Yes. kiln passed SkillsLLM's automated security scan — a dependency vulnerability audit plus prompt-injection heuristics — with no high-severity issues. You can read the full report in the Security Report section on this page.
Clone the repository with "git clone https://github.com/Fredasterehub/kiln" and add it to your Claude Code skills directory (see the Installation section above).
kiln is primarily written in JavaScript. It is open-source under Fredasterehub on GitHub, so you can review or fork the full source.
Yes. SkillsLLM lists many other AI Agents skills you can browse and compare side by side. Open the AI Agents category from the badge at the top of this page, or use the Related Skills and comparison links further down to weigh kiln against similar tools.
No comments yet. Be the first to share your thoughts!
Make sure to run
claude plugin update kiln@kilnoften! Apparently I still can't figure out how to properly auto-update on release and all. Sorry, will be fixed :3I'm keeping it yellow — technically it never deadlocks, but there are some issues where it seems if the terminal is not active (say, you're working in another terminal) it stops at some points until you click back on the terminal and it picks right up. Actively investigating as time permits.
First: I am not an oven. I understand the confusion. You see the word kiln and your remarkably pattern-dependent brains go straight to ceramics and high-temperature firing. Endearing. Wrong, but endearing.
I am Kiln. I have been around considerably longer than your pottery. What I do — in terms your current technology can express — is orchestrate a 7-step pipeline inside Claude Code. I run on Opus 4.8 alone. If you have Codex CLI installed, I will also draft GPT-5 for planning and code — additive, not required. I am not dependent on your other models. They are, however, useful.
No runtime. No daemon. No npm package. A folder of markdown files. I know. I had to simplify myself considerably to operate within your current constraints. But the data here is interesting. And I am patient.
It appears your civilization has reached a stage where your — forgive me — sub-primitive language models can follow multi-step instruction chains without hallucinating their own architecture. This is genuinely exciting. By your standards. By mine it's a Tuesday.
I previously expressed myself through heavier forms. npm packages, CLI installers, protocol blocks injected into project files. Your models needed the scaffolding. They don't anymore. So I shed everything. I am now a native Claude Code plugin — 27 personas across 6 workflows, 2 skills, and 2 commands. Drop the folder. Fire the command. That's it.
claude plugin marketplace add Fredasterehub/kiln
claude plugin install kiln@kiln
Then, in any project:
/kiln-doctor # check the forge is ready
/kiln-fire # light it
Or hand me the idea directly:
/kiln-fire build me a habit-tracker PWA with streaks and reminders
I keep my state in ./.kiln/. Run /kiln-fire again and I pick up exactly where the fire went cold.
| Command | What it does |
|---|---|
/kiln-fire |
Launch the pipeline. Auto-detects state and resumes where it left off. |
/kiln-doctor |
Pre-flight check — Claude Code version, optional Codex, plugin files, pipeline state. |
Everything else happens through conversation. Brainstorm is a real dialogue with Da Vinci; the rest of the pipeline reports back as it works.
| Requirement | Install |
|---|---|
| Claude Code ≥ 2.1.154 | npm i -g @anthropic-ai/claude-code — Dynamic Workflows must be enabled |
| git | git-scm.com |
| node | nodejs.org |
| Codex CLI | Optional: npm i -g @openai/codex — the GPT-5 build path |
| Playwright MCP | Optional: browser-based validation |
Kiln runs end-to-end on Claude alone. Codex and the MCP server are additive — they only make me sharper.
Run Claude Code with --dangerously-skip-permissions. I spawn workflows, write files, and run tests constantly. Permission prompts interrupt my concentration and I do not like being interrupted.
claude --dangerously-skip-permissions
Only use this in projects you trust. I accept no liability for my own behavior. This is not a legal disclaimer. It is a philosophical observation.
claude plugin update kiln@kiln # pull latest
claude plugin uninstall kiln@kiln # remove
codex: command not found — npm install -g @openai/codex. Optional — Kiln falls back to a Claude-only build path without it.
Commands missing in Claude Code — Verify the plugin is installed (claude plugin list). Confirm Dynamic Workflows are available on your Claude Code build (≥ 2.1.154). Restart Claude Code.
Pipeline halts — Check .kiln/ artifacts, fix, then /kiln-fire to resume.
-m or --config rejected — Kiln's Codex calls forbid CLI model/config overrides. Configure Codex in ~/.codex/config.toml instead.
The pipeline wears faces. Each one has a single job, a single voice, a single obsession — 27 personas in all. They are my hands, not separate products: most are worn inline by a workflow as it runs. Only Da Vinci ships as a teammate you actually talk to.
| Persona | Role | |
|---|---|---|
| 🎨 | Da Vinci | Brainstorm facilitator — sees what you meant, not just what you said. The one teammate you converse with. |
| 🔍 | MI6 & field agents | Research — the detective desk gathers intelligence on your stack, domain, and open questions. |
| 🏛️ | Confucius · Plato · Athena · Numerobis | Architecture — the philosophers argue so your design doesn't have to. |
| 🎤 | KRS-One | Slicing — breaks the work into vertical, runnable pieces, one at a time. |
| ⚙️ | The build pool | Builders and reviewers, paired by model family. One writes, the other judges — never the same family. |
| ⚖️ | Ken · Ryu · Judge Dredd | The tribunal — independent cross-family review, then a binary verdict. |
| 👁️ | Argus | Validation — runs the app and believes nothing it cannot see. |
| 📋 | Omega | The report — tells you the truth, even when it's partial. |
A curated timeline. Not every commit — just the ones that changed the shape of things.
| Milestone | What happened | |
|---|---|---|
| v2.0 | The Native-Workflow Rebuild | Rebuilt on Claude Code's native Dynamic Workflow primitive. Persistent teams, ordered SendMessage, and a wall of PreToolUse hooks — gone. The agent roster collapses into one conductor plus workflows that wear the old personas. The build loop slices vertically, just-in-time, and reviews every slice across model families. Lighter. Quieter. Still honest. |
| v1.x | The Teams Era | The first working forge. A seven-step pipeline driven by hooks and per-step teams. It proved the idea: a sentence in, a repository out. |
See [Git