ClawHub

Kimi2.5skill

v1.0.0

Troubleshoot and operate Kimi 2.5 / OpenClaw image understanding. Use when image recognition fails, OCR/images cannot be analyzed, `image` tool reports `Unkn...

⭐ 0· 52·0 current·0 all-time
by@jiangwill2023

Install

OpenClaw Prompt Flow

Install with OpenClaw

Best for remote or guided setup. Copy the exact prompt, then paste it into OpenClaw for jiangwill2023/kimi2-5skill.

Previewing Install & Setup.
Prompt PreviewInstall & Setup
Install the skill "Kimi2.5skill" (jiangwill2023/kimi2-5skill) from ClawHub.
Skill page: https://clawhub.ai/jiangwill2023/kimi2-5skill
Keep the work scoped to this skill only.
After install, inspect the skill metadata and help me finish setup.
Use only the metadata you can verify from ClawHub; do not invent missing requirements.
Ask before making any broader environment changes.

Command Line

CLI Commands

Use the direct CLI path if you want to install manually and keep every step visible.

OpenClaw CLI

Bare skill slug

openclaw skills install kimi2-5skill

ClawHub CLI

Package manager switcher

npx clawhub@latest install kimi2-5skill
Security Scan
VirusTotalVirusTotal
Benign
View report →
OpenClawOpenClaw
Benign
high confidence
✓
Purpose & Capability
Name/description match the instructions: the skill is a runbook for troubleshooting image/OCR failures and directs checks of OpenClaw runtime and agent-level models.json files, restart of gateway, and regression tests — all coherent with the stated purpose.
ℹ
Instruction Scope
Instructions explicitly tell the operator/agent to inspect and edit config files under ~/.openclaw and to restart the gateway and run the 'image' tool. That is expected for an ops playbook, but it does instruct reading and modifying user-specific files and performing privileged operations (restart/config edits) — these should be done under human supervision and with backups.
✓
Install Mechanism
No install spec and no code files — the skill is instruction-only, so nothing is written to disk or downloaded by the skill itself.
ℹ
Credentials
The skill does not request environment variables or credentials. However, the referenced models.json/provider blocks could contain provider credentials in practice; the playbook's advice to remove provider blocks may expose or alter those secrets, so operators should back up files and confirm credential handling before editing.
✓
Persistence & Privilege
always is false and the skill does not try to persist or modify other skills or global agent settings. Be cautious if the agent is allowed to invoke skills autonomously: following the playbook could involve automated edits and restarts if the agent acts without human oversight.
Assessment
This is an ops playbook (no code) and it is internally consistent with its purpose. Before using it: 1) run the troubleshooting steps manually first (openclaw status, a real image test) to confirm symptoms; 2) back up any models.json files before editing; 3) inspect provider blocks for embedded credentials and handle them carefully; 4) avoid letting the agent perform edits and gateway restarts autonomously — run changes under a human operator or with appropriate access controls; 5) verify the paths referenced (they include examples like /Users/admin-ai) match your environment. If you want the agent to perform any edits automatically, restrict that to trusted, privileged agents and log all changes.

Like a lobster shell, security has layers — review code before you run it.

latestvk974z171bbqtfy92rzdcc8nkc985kvj2
52downloads
0stars
1versions
Updated 2d ago
v1.0.0
MIT-0
@jiangwill2023
latest
Download

kimi2.5skill

处理 OpenClaw 中 Kimi 2.5 / 识图模型的排障、恢复、验证与运维收口。

快速流程

  1. 先做 runtime 实测，不要只看配置文件
  2. 再检查：
    • ~/.openclaw/openclaw.json
    • ~/.openclaw/agents/<agent>/agent/models.json
  3. 如果出现 Unknown model，优先怀疑 agent 级 models.json 的非法 provider 配置导致整份自定义模型注册表加载失败
  4. 修复后必须重启 gateway，并跑一张真实图片回归测试

参考文件

  • 修复流程：references/recovery-playbook.md
  • 本次事件：references/incident-2026-04-14.md

本轮已验证结论（2026-04-14）

本轮问题的真实根因是：

agent 级 models.json 中的自定义 codex provider 配置不合法，导致模型注册表加载异常，继而让 image runtime 报 Unknown model。

已验证修复点：

  • /Users/admin-ai/.openclaw/agents/qiang/agent/models.json
  • /Users/admin-ai/.openclaw/agents/sally-bot/agent/models.json

移除有问题的 codex provider 后，image 工具已成功返回图片描述，说明识图 runtime 恢复。

成功标准

以下同时满足才算修好：

  • 不再出现 Unknown model
  • image 工具返回正常图片描述 / OCR 结果
  • gateway 状态正常
  • 至少完成一张真实图片回归测试

收口口径

恢复后统一表述：

  • 识图已恢复可用
  • 根因是 agent 级 models.json 非法 provider 配置污染了 image runtime 注册表
  • 后续若再遇 Unknown model，优先检查 ~/.openclaw/agents/<agent>/agent/models.json

Comments

Loading comments...