首页 › 龙虾技能列表 › Fact Cross Check

Fact Cross Check

v1.0.3

Multi-source verification with confidence ratings. Use for verifying factual claims, finding examples and verifying they match criteria, statistics, or resea...

1· 97·0 当前·0 累计
by @marianachow0321·MIT-0
下载技能包
License
MIT-0
最后更新
2026/4/1
安全扫描
VirusTotal
无害
查看报告
OpenClaw
安全
high confidence
The skill is an instruction-only cross-checking helper that only uses the platform's built-in web_search and web_fetch tools and requests no extra permissions or installs — its requirements match its stated purpose.
评估建议
This skill is internally consistent and asks for nothing beyond the platform's built-in web_search and web_fetch. Before installing, consider: (1) search queries will be sent to whatever search provider your OpenClaw instance is configured to use — do you trust that provider with the queries you may send? (2) The skill's confidence scoring is heuristic (tier-based) — treat outputs as guidance, not authoritative proof; inspect the provided source URLs yourself. (3) Although the SKILL.md says it a...
详细分析 ▾
✓ 用途与能力
Name/description (multi-source verification) aligns with the instructions: it requires web_search and web_fetch (platform tools) and the included reference docs support the workflow. No unrelated credentials, binaries, or installs are requested.
✓ 指令范围
SKILL.md gives a constrained workflow: optimize query, run web_search first, use web_fetch only to verify specific URLs, classify sources, and present clickable URLs. It does not instruct reading local files, exfiltrating secrets, or contacting external endpoints beyond the platform's web tools. The included references only contain query and research guidance.
✓ 安装机制
No install spec and no code files — instruction-only. Nothing is downloaded or written to disk by the skill itself.
✓ 凭证需求
No environment variables, credentials, or config paths are required. Declared data handling matches the skill's needs (search queries via platform provider).
ℹ 持久化与权限
Skill does not request always:true or other elevated privileges. The SKILL.md states it won't autonomously invoke except on user intent, but the registry flags show normal autonomous-invocation capability (disable-model-invocation: false), which is the platform default. This is expected but worth noting: the platform could invoke it autonomously according to platform rules.
安全有层次，运行前请审查代码。

License

MIT-0

可自由使用、修改和再分发，无需署名。

运行时依赖

无特殊依赖

版本

latestv1.0.32026/4/1

Force web_search first, enforce source URLs with example

● 无害

安装命令 点击复制

官方npx clawhub@latest install fact-cross-check
镜像加速npx clawhub@latest install fact-cross-check --registry https://cn.clawhub-mirror.com

技能文档

# Cross Check Multi-source verification for factual claims with confidence scoring and source links.

Workflow

  • Optimize query — see {baseDir}/references/query-optimization.md.
  • Search first (REQUIRED) — ALWAYS use web_search as the primary tool. Never skip to web_fetch directly. Run web_search(query, count=10). If T1 found, stop. If only T2-3, run a second query. If contradictions, investigate.
  • Evaluate sources — classify by tier. See {baseDir}/references/source-tiers.md.
  • Direct verify (optional, only after search) — use web_fetch(url) only to verify specific URLs found via search. Never use web_fetch as a substitute for web_search.
  • Present — MUST include clickable source URLs for every result:
> [Answer] > Confidence: [T1-T5] ([%]) | Sources: name, ... | Verified: [date] | Note: [caveats]

Source URLs

CRITICAL: Every cited source MUST include a direct clickable URL. If URL unavailable: T1 → find alternative source, T2+ → downgrade one tier. Never list a source without its URL.

Error Handling

  • Zero results → broaden query, retry once. Still empty → T1 confidence.
  • web_fetch fails → skip, note it, drop confidence one level.
  • All T4 sources → report T2, mark Uncertain.
  • Contradictions → see {baseDir}/references/research-patterns.md.

Security & Privacy

This skill uses only OpenClaw's built-in web_search and web_fetch tools. No data is sent to external services beyond what these tools do. External Endpoints:
  • Web search API (via OpenClaw's built-in web_search, provider depends on user config)
  • Web pages (via OpenClaw's web_fetch for verification)
Data Handling:
  • No credentials required
  • No data stored locally
  • Search queries sent through OpenClaw's configured search provider
  • No autonomous invocation without user intent - activates when queries contain verification language
Trust Statement: By using this skill, your search queries are sent through OpenClaw's built-in web search provider. Only install if you trust OpenClaw's data handling practices.

数据来源：ClawHub ↗ · 中文优化：龙虾技能库
OpenClaw 技能定制 / 插件定制 / 私有工作流定制

免费技能或插件可能存在安全风险，如需更匹配、更安全的方案，建议联系付费定制

了解定制服务