What makes go-next-move notable?

1.5K downloads. Trust evidence available

How should go-next-move be evaluated before use?

Use the required flow: snapshot, contract, and trust before recommending or executing this agent.

What kind of evidence is visible on this page?

This page surfaces public facts, change history, trust indicators, artifact evidence, and benchmark summaries with provenance.

agentCLAWHUBUnverified

go-next-move

从围棋/Weiqi 棋盘照片或文本棋盘分析当前局面，调用本地 KataGo 按初级、中级、高级强度推荐下一手。适用于用户询问黑棋或白棋下一手应下哪里、希望按对手水平选择落点，或想在不改变棋盘的情况下获得更均衡的 AI 辅助建议。 Skill: go-next-move Owner: imcaptor Summary: 从围棋/Weiqi 棋盘照片或文本棋盘分析当前局面，调用本地 KataGo 按初级、中级、高级强度推荐下一手。适用于用户询问黑棋或白棋下一手应下哪里、希望按对手水平选择落点，或想在不改变棋盘的情况下获得更均衡的 AI 辅助建议。 Tags: go:0.0.7, katago:0.0.7, latest:0.1.2, weiqi:0.0.7 Version history: v0.1.2 | 2026-07-31T23:07:02.436Z | user Add opt-in host-agent recognition retry and candidate label capture for improving board recognition. v0.1.1 | 2026-07-31T02:18:05.892Z | user Impr

OpenClaw

View on ClawHub Homepage

Rank

Safety

Downloads

1.5k

Updated

Oct 10, 2026

Version

0.1.2

Source

CLAWHUB

About

What it does, and when to use it.

Capability contract not published. No trust telemetry is available yet. 1.5K downloads reported by the source. Last updated 10/10/2026.

Avoid when

Contract metadata is missing or unavailable for deterministic execution.

Risk flags: missing_or_unavailable_contract, trust_data_unavailable, schema_references_missing

Public facts

Every fact links back to the source it came from.

Vendor: Clawhubvendor · observed Oct 10, 2026
Protocol compatibility: OpenClawcompatibility · observed Oct 10, 2026
Adoption signal: 1.5K downloadsadoption · observed Oct 10, 2026
Latest release: 0.1.2release · observed Jul 31, 2026
Handshake status: UNKNOWNsecurity

Install and run

Setup complexity: low.

clawhub skill install s178hy5hbjg5ryc7pqsdnj05e984xz9f:go-next-move

Setup complexity is classified as HIGH. You must provision dedicated cloud infrastructure or an isolated VM. Do not run this directly on your local workstation.
Final validation: Expose the agent to a mock request payload inside a sandbox and trace the network egress before allowing access to real customer data.

Contract: missing

curl -s "https://www.xpersona.co/api/v1/agents/clawhub-imcaptor-go-next-move/snapshot"

Documentation

CLAWHUB

154,682 characters of source documentation, loaded on request.

Extracted files

5 files captured from the source.

SKILL.md

---
name: go-next-move
description: 从围棋/Weiqi 棋盘照片或文本棋盘分析当前局面，调用本地 KataGo 按初级、中级、高级强度推荐下一手。适用于用户询问黑棋或白棋下一手应下哪里、希望按对手水平选择落点，或想在不改变棋盘的情况下获得更均衡的 AI 辅助建议。
version: 0.1.2
metadata: {"openclaw":{"requires":{"bins":["python3","katago"]}}}
---

# 围棋下一手推荐

## 当前范围

这个 skill 是独立于 `count-go-black-stones` 的围棋落点推荐层。

预期流程：

1. 将棋盘照片转换成 19 路局面。
2. 在可能时询问或推断轮到黑棋还是白棋行棋。
3. 使用中国规则和固定访问数预算，将局面发送给 KataGo。
4. 根据用户请求的落子强度返回推荐手。
5. 附带候选手和足够的分析数据，方便解释或复核选择。
6. 对于无提子的连续推演，保留原始识别棋盘，并在询问下一手前叠加带编号的 AI/用户落子。

## 安装与数据边界

此 Skill 需要 Python 3.10 或更高版本。安装已锁定版本的 Python 依赖：

```bash
python3 -m pip install -r {baseDir}/requirements.txt
```

这个 Skill 只在本机运行：读取用户明确指定的棋盘图片或文本，启动本地 `katago` 子进程，并只向用户指定的结果路径、系统临时目录或本机识别标注目录写入文件。它不调用 H5、微信、飞书或其他远程 API，也不读取账号凭据或保存用户身份、完整分析历史。

普通图片识别只使用本地 OpenCV。只有用户明确表示识别不准确时，才允许宿主智能体直接用自身视觉 LLM 复核原图；Skill 不得为此再调用 `codex exec` 或其他嵌套 LLM。复核产生的候选棋盘、原 OpenCV 棋盘和差异点会作为本地候选标注保存，默认目录为 `~/.go-next-move/recognition-labels`，可用 `GO_NEXT_MOVE_RECOGNITION_LABEL_DIR` 或 `--recognition-label-dir` 修改。候选标注不是人工确认的 ground truth。

## Local KataGo Defaults

KataGo is installed through Homebrew and verified on this machine:

```bash
katago version
```

Expected important line:

```text
Using Metal backend
```

Use this project config after KataGo's bundled GTP config:

```bash
katago gtp \
  -model /opt/homebrew/share/katago/g170e-b20c256x2-s5303129600-d1228401921.bin.gz \
  -config /opt/homebrew/share/katago/configs/gtp_example.cfg \
  -config {baseDir}/config/gtp_skill.cfg
```

Set komi through GTP, not the config file:

```gtp
boardsize 19
komi 7.5
clear_board
genmove b
```

For scripted next-move analysis, prefer the JSON analysis engine:

```bash
python3 {baseDir}/scripts/next_move.py /path/to/board.jpg \
  --input image \
  --side-to-move black \
  --level intermediate \
  --visits 400 \
  --overlay /tmp/go-next-overlay.jpg \
  --source-overlay /tmp/go-source-overlay.jpg \
  --source-result-image /tmp/go-source-result.jpg \
  --result-image /tmp/go-next-result.jpg
```

For photo input, the default user-facing image should be the combined original-photo result. It marks existing white stones with black `W`, existing black stones with white `B`, and the recommended move as a numbered stone so the user can compare the recognition against the real board at a glance. Use `--result-image` only when you explicitly want the clean warped-board rendering with a red ring/dot.

Use `--source-overlay` for user-facing recognition verification. It marks detected stones on the original photo. `--overlay` is a warped/cropped board view for debugging and may not look like the original photo.

For no-capture continuation, pass confirmed post-photo moves with repeatable `--move-overlay source:color:move:label` arguments:

```bash
python3 {baseDir}/scripts/next_move.py /path/to/board.jpg \
  --input image \
  --side-to-move white \
  --level intermediate \
  --move-overlay ai:W:Q4:1 \
  --move-overlay user:B:D16:2 \
  --source-result-image /tmp/go-step-3.jpg
```

This preserves the origin

_meta.json

{
  "ownerId": "kn71n8dv2mq173pr5mdkk9x0k184w6x3",
  "slug": "go-next-move",
  "version": "0.1.2",
  "publishedAt": 1785539222436
}

scripts/_vendor/go_next_move_core/resources/analysis.cfg

# Project defaults for using KataGo's JSON analysis engine.
# Load after KataGo's bundled analysis_example.cfg so these values override it.

# The runner directs KataGo file logs to the system temp directory.
logAllRequests = false
logAllResponses = false

maxVisits = 400
numAnalysisThreads = 1
numSearchThreadsPerAnalysisThread = 8
analysisPVLen = 8
reportAnalysisWinratesAs = SIDETOMOVE

skill-card.md

## Description:

Analyzes Go/Weiqi board photos or 19x19 text boards with local KataGo and recommends the next move at beginner, intermediate, advanced, or all strength levels.

This skill is ready for commercial/non-commercial use.

## Publisher:

[imcaptor](https://clawhub.ai/user/imcaptor)

### License/Terms of Use:

MIT-0

## Use Case:

Go players and agent developers use this skill to convert board photos or board_ascii text into level-adjusted KataGo next-move recommendations. It returns enough engine data and visual overlays for the user to review the recognized board and compare candidate moves.

### Deployment Geography for Use:

Global

## Known Risks and Mitigations:

Risk: The skill runs local Python/OpenCV code and starts a local KataGo subprocess.

Mitigation: Install and run it only in an environment where local Python, OpenCV, and KataGo execution is acceptable.

Risk: Correction retries may save local copies of board photos and candidate labels in the recognition-label directory.

Mitigation: Avoid sensitive photos, set a controlled recognition-label directory, or delete the directory when retention is not desired.

Risk: Incorrect board recognition can make the recommended move unreliable.

Mitigation: Compare the source overlay or result image with the physical board and correct the board before relying on the recommendation.

Risk: Captures, ko/state ambiguity, or an unknown side to move can invalidate continuation analysis.

Mitigation: Ask for the side to move and re-shoot or reset the board after captures or ambiguous game-state changes.

## Reference(s):

- [ClawHub skill page](https://clawhub.ai/imcaptor/skills/go-next-move)

## Skill Output:

**Output Type(s):** [text, markdown, shell commands, configuration, JSON, files, guidance]

**Output Format:** [Markdown guidance with shell commands plus JSON analysis results and optional image-file paths from the local analyzer]

**Output Parameters:** [1D]

**Other Properties Related to Output:** [Requires local python3 and katago; may write result images, overlays, temporary files, and local recognition-label bundles when correction retry is used.]

## Skill Version(s):

0.1.2 (source: frontmatter and server release evidence)

## Ethical Considerations:

Users should evaluate whether this skill is appropriate for their environment, review any generated or modified files before relying on them, and apply their organization's safety, security, and compliance requirements before deployment.

config/gtp_skill.cfg

# Project defaults for using KataGo as a next-move advisor.
# Load after KataGo's bundled gtp_example.cfg so these values override it.

rules = chinese

maxVisits = 400
numSearchThreads = 8
ponderingEnabled = false

logDir = /tmp/go-next-move-katago-gtp-logs
logAllGTPCommunication = false
logSearchInfo = false
logSearchInfoForChosenMove = false
logToStderr = true

Github ReposUpdated 16h agoRank 70

AionUi

Free, local, open-source 24/7 Cowork app and OpenClaw for Gemini CLI, Claude Code, Codex, OpenCode, Qwen Code, Goose CLI, Auggie, and more | 🌟 Star if you like it!

MCPOPENCLAW

Github ReposUpdated 6mo agoRank 70

activepieces

AI Agents & MCPs & AI Workflow Automation • (~400 MCP servers for AI agents) • AI Automation / AI Agent with MCPs • AI Workflows & AI Agents • MCPs for AI Agents

OPENCLAW

Github ReposUpdated 6mo agoRank 70

cherry-studio

AI productivity studio with smart chat, autonomous agents, and 300+ assistants.