在本地磁盘以分类纯 Markdown 文件保存需要长期留存的事实,与智能体内置记忆并存。
数据分析
skilled-openclaw-advisor
试用Query the local OpenClaw docs index for accurate answers about configuration, features, CLI commands, channels, providers, plugins, cron, sessions, agents, protocol, and troubleshooting. Faster and more accurate than relying on training data for OpenClaw specifics. Zero API calls, sub-10ms queries. Useful for: openclaw, configure, gateway, channel, cron, provider, plugin, session, heartbeat, protocol, skill, model, agent questions.
它能做什么
Local FTS5 index of all OpenClaw documentation. No external API calls. Sub-10ms queries.
技能文档
OpenClaw Docs Intelligence
Local FTS5 index of all OpenClaw documentation. No external API calls. Sub-10ms queries.
What This Skill Does
This skill bundles three Python scripts and queries a local SQLite FTS5 index
built from the OpenClaw docs installed on this machine
(~/.npm-global/lib/node_modules/openclaw/docs or equivalent). It reads only
the local docs directory and writes only to ~/.openclaw/skills-data/skilled-openclaw-advisor/.
The query and build scripts make no network requests. The update script optionally sends a local notification via the openclaw message CLI when doc changes are detected (no direct network calls). The scripts do not access credentials or config secrets,
and does not read openclaw.json at runtime (config keys listed below are optional
overrides that the user may set — they are not required).
First-Time Setup
Build the index after installing the skill:
python3 $SKILL_DIR/scripts/build_index.py
This scans the local OpenClaw docs and creates a SQLite index at
~/.openclaw/skills-data/skilled-openclaw-advisor/index.db.
Query
# Compact output for agent use
python3 $SKILL_DIR/scripts/query_index.py --query "YOUR QUESTION" --mode agent
# Human-readable
python3 $SKILL_DIR/scripts/query_index.py --query "YOUR QUESTION"
# Full detail with online link
python3 $SKILL_DIR/scripts/query_index.py --query "YOUR QUESTION" --verbosity detailed
# Chinese results
python3 $SKILL_DIR/scripts/query_index.py --query "YOUR QUESTION" --lang zh-CN
# Index status
python3 $SKILL_DIR/scripts/query_index.py --status
Index Updates
# Incremental update (checks for doc changes)
python3 $SKILL_DIR/scripts/update_index.py
# Force full re-index
python3 $SKILL_DIR/scripts/build_index.py --force
When to Use
- OpenClaw configuration questions
- CLI commands and options
- Channel, provider, cron, session, agent setup
- Troubleshooting OpenClaw errors
When Not to Use
- General coding questions unrelated to OpenClaw
- Questions about third-party services
- If index is not built yet (run
build_index.pyfirst)
Response Modes
| Mode | Use Case | Description |
|---|---|---|
agent | Default agent queries | Ultra-compact, minimal tokens |
human + standard | Human questions | Clear explanation + key points |
human + detailed | Deep dives | Full excerpt + examples |
Optional Config
Settings in skills.entries.openclaw-docs.config in openclaw.json (all optional):
| Key | Default | Description |
|---|---|---|
defaultLang | "en" | Language: en or zh-CN |
defaultVerbosity | "standard" | Output: concise, standard, detailed |
maxResults | 5 | Max results returned |
fallbackToEnglish | true | Fall back to English if no results in requested language |
includeOnlineUrl | true | Include https://docs.openclaw.ai/... links in detailed output |
相关技能
通过托管的 OAuth GraphQL 接口查询与管理 Linear 的 issue、项目、团队、周期、标签和评论。
通过 OAuth 认证网关管理 Stripe 客户、订阅、发票、产品、价格和支付。
用 Python 自适应抓取网页,默认绕过反爬保护,支持从单次请求到大规模并发爬取。
诊断生产力系统反复失效的根因,给出最小干预——容量测算、瓶颈定位、可靠的本地记录。
以 AI 机器人身份加入视频会议,提供语音、虚拟形象与屏幕共享四种模式。
seanford 的更多技能
浏览全部技能通过 Microsoft Graph 接入 Outlook,读取、发送、管理邮件、文件夹、日历事件和联系人,OAuth 由平台托管。
通过托管网关调用第三方 API,免去自行管理 OAuth 和密钥的负担。
通过托管 OAuth 代理读写并管理 Gmail 邮件、会话、标签和草稿。
Multi-provider web search and URL content extraction with intelligent auto-routing, freshness filters, locale awareness, caching, and privacy options across...
通过 acpx 把编码任务交给 Codex、Claude Code、Pi 等 ACP 兼容代理,支持常驻会话与并行执行。
Expert web fetching and crawling using Scrapling. Use for any web_fetch task, JS-rendered pages, anti-bot sites, bulk URL fetching, or site crawling. Preferr...