https://github.com/tivaliy/skiller
Declarative, human-in-the-loop workflow runner for VS Code chat - branching YAML playbooks that orchestrate LLMs and MCP tools
https://github.com/tivaliy/skiller
ai ai-orchestration ai-skills ai-workflows chat-participant
Last synced: 15 days ago
JSON representation
Declarative, human-in-the-loop workflow runner for VS Code chat - branching YAML playbooks that orchestrate LLMs and MCP tools
- Host: GitHub
- URL: https://github.com/tivaliy/skiller
- Owner: tivaliy
- License: mit
- Created: 2026-06-15T15:44:55.000Z (about 1 month ago)
- Default Branch: main
- Last Pushed: 2026-06-17T19:45:41.000Z (about 1 month ago)
- Last Synced: 2026-06-17T21:22:01.974Z (about 1 month ago)
- Topics: ai, ai-orchestration, ai-skills, ai-workflows, chat-participant
- Language: TypeScript
- Homepage: https://marketplace.visualstudio.com/items?itemName=tivaliy.skiller
- Size: 2.72 MB
- Stars: 1
- Watchers: 0
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- Changelog: CHANGELOG.md
- License: LICENSE
Awesome Lists containing this project
README
# Skiller
**Declarative, human-in-the-loop workflow runner for VS Code chat.**
Write branching YAML playbooks that orchestrate your language model and MCP tools — with review and input steps — and run them from the `@skiller` chat participant.
[](https://marketplace.visualstudio.com/items?itemName=tivaliy.skiller)
[](LICENSE)
### 📖 Full documentation →
---
## What is Skiller?
Skiller runs **skills** — declarative, multi-step workflows defined in YAML and Markdown. Each skill is a playbook: a sequence of steps that call a language model, invoke your [MCP](https://modelcontextprotocol.io) tools, and pause for your input or approval exactly where you tell them to.
It is deliberately **not** a free-form agent and **not** an "Agent Skills" auto-loader:
| | Free-form agent | **Skiller** |
| --- | --- | --- |
| Control flow | Model decides what to do next | **You** define the steps; the model fills them in |
| Side effects | Can act on its own | Runs **only** the steps you wrote, pausing for confirmation where you ask |
| Reproducibility | Varies run to run | Same playbook, same shape every time |
| Branching | Implicit | Explicit `confirmation` steps with `goto` jumps |
If you want predictable, reviewable automation that still leverages an LLM — not an autonomous agent — Skiller is for you. As a skill runs, you can watch it as a **live execution graph**, with branches and loops lighting up as they fire.
## Requirements
- **VS Code 1.93+**
- A **chat language model provider** (e.g. GitHub Copilot, or any provider exposing VS Code's Language Model API)
- **MCP servers** configured in VS Code (optional — only needed for tool integrations)
## Quick start
### Install
**From the Marketplace:**
```text
ext install tivaliy.skiller
```
**From source:**
```bash
npm install
npm run package
code --install-extension skiller-*.vsix
```
### Run your first skill
Open the Chat view and talk to `@skiller`:
```text
@skiller /help Show available commands
@skiller /skills List discovered skills
@skiller /skill greeter Run the built-in "greeter" example skill
```
A message that is **not** a slash command and **not** a skill gets a short hint — Skiller does not do free-form chat.
See the [Getting Started tutorial](https://tivaliy.github.io/skiller/getting-started/introduction/) to install, run a skill, and author your own.
## Learn more
The docs are the single source of truth. Jump to what you need:
- **[Run a built-in skill](https://tivaliy.github.io/skiller/getting-started/run-a-bundled-skill/)** and **[Write your first skill](https://tivaliy.github.io/skiller/getting-started/write-your-first-skill/)** — guided tutorials.
- **Guides** — how-to for [tool/MCP skills](https://tivaliy.github.io/skiller/guides/tool-mcp-skill/), [branching & looping](https://tivaliy.github.io/skiller/guides/branching-looping/), [debugging](https://tivaliy.github.io/skiller/guides/debugging/), and [models](https://tivaliy.github.io/skiller/guides/models/).
- **[Skills & discovery](https://tivaliy.github.io/skiller/concepts/skills-and-discovery/)**, **[Step types & state](https://tivaliy.github.io/skiller/concepts/step-types/)**, **[Templating with Liquid](https://tivaliy.github.io/skiller/concepts/templating/)**, and **[The execution graph](https://tivaliy.github.io/skiller/concepts/execution-graph/)** — how it all works.
- **[skill.yaml manifest](https://tivaliy.github.io/skiller/reference/skill-yaml/)** and **[step-file frontmatter](https://tivaliy.github.io/skiller/reference/step-frontmatter/)** — every field.
- **[Commands](https://tivaliy.github.io/skiller/reference/commands/)** — all `@skiller` slash commands.
- **[Settings](https://tivaliy.github.io/skiller/reference/settings/)** — every `skiller.*` setting and its default.
- **[Built-in tools](https://tivaliy.github.io/skiller/reference/built-in-tools/)** — `skiller_createFile` and `skiller_replaceInFile`.
## Contributing
Issues and pull requests are welcome. See **[Contributing & development](https://tivaliy.github.io/skiller/contributing/)** for build, test, and PR guidelines, and for running the Extension Development Host.
## License
Licensed under the [MIT License](LICENSE).