← 返回 Skills 市场
Fact Checker
作者
cliffyan28
· GitHub ↗
· v1.0.1
· MIT-0
432
总下载
1
收藏
2
当前安装
2
版本数
在 OpenClaw 中安装
/install openclaw-fact-checker
功能描述
Fact-check news articles, social media posts, images, and videos. Use when verifying claims, detecting deepfakes or AI-generated content, identifying out-of-...
安全使用建议
This instruction-only skill appears internally consistent with a fact‑checking purpose. It will fetch and analyze user-provided URLs, images, and videos (and may download media to /tmp briefly). No environment variables are required, but you can optionally provide API keys (Google Fact Check, Brave Search, OpenAI) to enable extra features. Before installing, consider: (1) ensure you are comfortable allowing the skill to perform web searches and download user-supplied media for analysis, (2) optional tools (ffmpeg, exiftool, c2patool, whisper) increase analysis depth but are not required, and (3) if you lack a configured web search, the skill will not perform text verification until you set one up (it suggests Brave Search).
功能分析
Type: OpenClaw Skill
Name: openclaw-fact-checker
Version: 1.0.1
The skill is a comprehensive multimodal fact-checking tool designed to verify text, images, and videos using web search, vision capabilities, and standard CLI utilities like ffmpeg and exiftool. While it executes shell commands and suggests a configuration update to AGENTS.md for persistence, these actions are transparently aligned with its stated purpose and follow standard OpenClaw patterns. No evidence of malicious intent, data exfiltration, or unauthorized remote control was found across the pipeline files (text_pipeline.md, image_pipeline.md, video_pipeline.md).
能力评估
Purpose & Capability
Name/description (multimodal fact checking) aligns with the instructions and references: text/image/video pipelines, reverse image search, EXIF/C2PA checks, ffmpeg/whisper usage, and optional Google/Brave/OpenAI APIs. No unrelated secrets or binaries are required.
Instruction Scope
SKILL.md gives detailed, narrowly scoped runtime steps that stay within fact‑checking (web searches, WebFetch, reverse image search, metadata extraction, ffmpeg keyframe extraction, transcription). It requires access to arbitrary URLs and user‑provided media (expected), writes temporary files under /tmp, and conditionally calls optional CLI tools. Note: the text pipeline mandates having web search available and explicitly refuses to proceed without it for text verification — this is an operational dependency rather than a security incoherence.
Install Mechanism
Instruction-only skill with no install spec and no downloaded code. Optional tooling is invoked only if present on PATH; recommended install commands are typical package manager invocations. No external arbitrary archive downloads or custom install scripts are present.
Credentials
No required environment variables or credentials. Optional env vars are appropriate and proportionate to functionality (GOOGLE_FACTCHECK_API_KEY for the Fact Check API, BRAVE_SEARCH_API_KEY for web search, OPENAI_API_KEY for Whisper API). No unrelated secrets or multiple unrelated credentials requested.
Persistence & Privilege
Skill is not forced‑always, does not request elevated or cross‑skill configuration changes, and has default autonomous-invocation settings. It does not ask to persist credentials or modify other skills.
如何使用
- 确保已安装 OpenClaw（本地或 Docker 部署）
- 在对话框中输入安装命令：
/install openclaw-fact-checker - 安装完成后，直接呼叫该 Skill 的名称或使用
/openclaw-fact-checker触发 - 根据 Skill 的参数说明提供必要输入，即可获得结构化输出
版本历史
v1.0.1
- Updated environment variable and binary dependencies to "optional" rather than "required".
- Added new optional binaries: whisper.
- Added new optional environment variables: OPENAI_API_KEY and BRAVE_SEARCH_API_KEY.
- Clarified setup instructions and dependency handling.
- No changes detected to core logic or execution pipelines.
v1.0.0
New skill version adds multimodal (text, image, video) fact-checking, multilingual support, and structured pipelines:
- Supports fact-checking news articles, social posts, images, and videos, including deepfake/AI detection.
- Introduces distinct pipelines for text, image, and video inputs, with language and modality auto-detection.
- Enforces report language to match user input, with clear label translation rules.
- Adds detailed confidence scoring, mandatory source citation, and strict objective/anti-hallucination guidelines.
- Removes previous script-based pipeline; now uses structured reference files for all modalities.
- Requires only specific CLI tools (ffmpeg, exiftool, c2patool) and disables external package dependencies.
元数据
常见问题
Fact Checker 是什么？
Fact-check news articles, social media posts, images, and videos. Use when verifying claims, detecting deepfakes or AI-generated content, identifying out-of-... 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件，目前累计下载 432 次。
如何安装 Fact Checker？
在 OpenClaw 或 Claude Code 对话框中运行命令「/install openclaw-fact-checker」即可一键安装，无需额外配置。
Fact Checker 是免费的吗？
是的，Fact Checker 完全免费，采用 MIT-0 许可证，可自由下载、安装和使用。
Fact Checker 支持哪些平台？
Fact Checker 跨平台运行，可在任意部署了 OpenClaw / Claude Code 的环境中使用（cross-platform）。
谁开发了 Fact Checker？
由 cliffyan28（@cliffyan28）开发并维护，当前版本 v1.0.1。
推荐 Skills