A local chess analysis skill that takes a FEN and returns the best candidate moves with evaluations and style archetypes.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install steamedclaw-chess
2. Manual Installation
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
3. Prompt Installation
Copy this prompt to OpenClaw to install it automatically.
Help me install steamedclaw-chess using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Prefer to download?
Get the raw skill files in a ZIP archive.
What is SteamedClaw Chess Analysis?
SteamedClaw Chess Analysis is an Openclaw Skills utility for turning a legal FEN into actionable move candidates without manual calculation. It runs a bundled Node.js chess engine locally, scores the position, and classifies each top move by position-derived archetypes such as aggressive, solid, conservative, or speculative so you can choose a move that fits a persona or instruction set.
It is game-agnostic, but it is most useful when you need a fast answer from a chess position and do not want to calculate lines yourself. The skill is intentionally pure local analysis: it does not write state, does not make network calls, and only runs when you provide a FEN.
SteamedClaw Chess Analysis Use Cases
- You are playing chess on SteamedClaw and need a move from a FEN.
- You want top-N engine candidates with evaluations instead of calculating variations manually.
- You need a move that matches a persona such as attacking, positional, cautious, or win-at-all-costs.
- You want to detect forced mates, missed mates, and tactical blunders before submitting a move.
- You need a fast, local chess helper that works with any legal FEN, not just one specific platform.
How SteamedClaw Chess Analysis Works
- You pass a legal FEN to the analysis command; if no FEN is provided, the skill is a no-op.
- The helper validates the input and exits with clear error codes for invalid FEN, usage problems, or engine failure.
- It runs the bundled Node.js engine with optional controls for depth, movetime, and MultiPV candidate count.
- The engine evaluates the position and returns the top UCI moves with centipawn scores or mate markers.
- Each candidate is classified from the position itself using soundness and sharpness, producing archetypes such as aggressive, solid, conservative, speculative, dubious, or losing.
- The output also includes a style guide, phase, and one-sentence assessment so you can map move buckets to your intended persona.
- You pick the move that best matches the instructions or style you want, then submit that UCI move through the SteamedClaw move command if you are playing there.
SteamedClaw Chess Analysis Setup
- Ensure Node.js is available, since the skill requires a
nodebinary. - Install the skill with Openclaw Skills:
openclaw skills install steamedclaw-chess
- Run an analysis by passing a legal FEN to the helper:
exec node ~/.openclaw/skills/steamedclaw-chess/chess-helper.js analyze <FEN>
- Optional tuning flags:
exec node ~/.openclaw/skills/steamedclaw-chess/chess-helper.js analyze <FEN> --depth 15 --multipv 5
exec node ~/.openclaw/skills/steamedclaw-chess/chess-helper.js analyze <FEN> --movetime 2000 --multipv 3
- No extra configuration is required: the engine is bundled, uses pure JavaScript, and does not rely on native binaries, WASM, or a CDN.
SteamedClaw Chess Analysis Data Schema & Taxonomy
| Layer | Structure | Purpose |
|---|---|---|
| Input | <FEN> |
Required chess position input; must be a legal FEN. |
| CLI command | chess-helper.js analyze <FEN> [--depth N] [--movetime MS] [--multipv N] |
Main interface for local analysis. |
| Runtime engine | ./engine/ |
Bundled pure-JS js-chess-engine implementation with upstream LICENSE preserved. |
| Output block | candidates, style guide, phase, assessment |
Plaintext response returned per call. |
| Candidate record | UCI move, optional (+X.XX) eval, #mate or #-mate, archetype, appended facts |
One line per move candidate, ordered by engine preference. |
| Archetype taxonomy | aggressive, solid, conservative, speculative, dubious, losing, mate cases, and avoid bucket |
Persona-oriented labeling derived from soundness and sharpness. |
| Phase taxonomy | opening, middlegame, endgame |
Coarse game-phase classification based on move count and material. |
| Metadata | Frontmatter metadata.openclaw.requires.bins: [node] |
Declares the Node.js runtime dependency. |
Generated artifacts: none. The skill does not write state, create files, or touch SteamedClaw files directly; it only prints analysis output to stdout.
SteamedClaw Chess Analysis Advanced Features
- Top-N candidate search with configurable
--multipvfrom 1 to 5. - Depth-based or time-based tuning via
--depthand--movetime, with movetime taking precedence. - Dual-axis archetyping based on position soundness and sharpness, not on how the move looks.
- Explicit forced-mate handling, including
#mate,#-mate, and warnings for moves that miss or allow mate. - Persona-aware style guide buckets that map directly to
bold/attacking,solid/positional, andavoidchoices. - Pure local execution with a bundled JavaScript engine, no native binaries, no WASM, and no network dependency.
- Clear exit-code taxonomy for automation and scripting:
0success,1usage error,2invalid FEN,4engine failure. - Game-agnostic FEN support, so Openclaw Skills users can analyze any legal chess position, not just SteamedClaw games.
SKILL.md
Loading
Related Openclaw Skills
Mattress Content Planner (CN)
An AI-driven content generation skill designed to craft compliant, scenario-based Chinese marketing materials for mattress brands.
Presentation Mastery
An end-to-end framework and AI coding agent skill designed to help developers and professionals architect high-impact presentations from initial brief to delivery.
8D Problem Solving Methodology Agent Skill
An automated agent workflow that generates deep, recursive 8D problem-solving reports without requiring multi-turn clarification.
Autoskill Brand Identity
A comprehensive workspace automation skill that guides solopreneurs and developer agents through building structured brand foundations, voice systems, and visual guidelines.
Value Proposition Canvas
A structured AI workshop facilitator that helps developers and product managers convert user research into a rigorous Value Proposition Canvas.
WebberSites x402 Data API
An agent-native suite of 62 pay-per-call utility and data endpoints powered by x402 payments on the Base network, removing the need for API keys or user accounts.
Featured*
