← Back to Skills Marketplace

Truth Seeking Fact Check

Name: Truth Seeking Fact Check
Author: tangtaozhanshen

by tangtaozhanshen · GitHub ↗ · v1.5.0 · MIT-0

cross-platform ⚠ suspicious

432

Downloads

Stars

Active Installs

Versions

Install in OpenClaw

/install truth-seeking-fact-check

Description

对信息进行权威多源事实核查，判定真伪并支持链上存证与用户反馈优化。

Usage Guidance

Things to check before installing or enabling this skill: - Privacy vs network use: The README/ SKILL.md promises “100% local” processing, but the code supports optional network lookups (Brave Search via requests) when you configure a datasource. If you need guaranteed offline operation, do not configure external datasources and review the code path that calls requests. - Scheduler behavior: If you enable the scheduler, it spawns a background thread that will call check_text periodically and may call a user-provided callback. Only enable scheduler in environments where background tasks and periodic network access are acceptable. - Configuration surface: The BraveSearch data source requires an API key in the skill config (not an env var). Confirm where you will store that key and whether you trust the skill/config to use it only for the intended searches. - Dependencies & install: There is no explicit install spec; a requirements.txt lists some libraries (beautifulsoup4, lxml, nltk) that aren’t referenced in the shipped code. Expect to manually review/install dependencies or run the skill in a virtual environment to avoid unnecessary packages on your system. - Metadata drift: version fields differ across _meta.json, openclaw.plugin.json, and manifest; this suggests sloppy packaging — review the code directly (you have it) and run it in an isolated environment before trusting it. - Audit for sensitive file access: The compliance module will attempt to read sensitive_words.txt from the module directory if present. If you have strict file‑access policies, inspect that file or remove it. Recommendation: If you plan to use this skill, audit and run it in a sandboxed environment first, avoid configuring external datasources unless you trust them, and only enable the scheduler after confirming desired behavior. The inconsistencies appear to be sloppy engineering rather than clearly malicious, but they increase risk in untrusted deployments.

Capability Analysis

Type: OpenClaw Skill Name: truth-seeking-fact-check Version: 1.5.0 The 'Truth (求真)' skill bundle is a functional fact-checking tool designed to identify AI hallucinations and verify content authenticity. The code implements text preprocessing, multi-dimensional scoring, and a scheduler for periodic checks (scheduler.py), all of which align with the stated purpose in SKILL.md. Network activity is limited to the Brave Search API for fact retrieval (datasource.py) and local parsing of blockchain URLs (blockchain.py). No evidence of data exfiltration, malicious execution, or harmful prompt injection was found.

Capability Assessment

ℹ Purpose & Capability

The code implements a local fact‑checking engine, sentence processing, scoring, batch support, a lightweight blockchain check, and an optional scheduler — all consistent with the skill description. However the DataSourceManager includes a BraveSearchDataSource that will perform outbound HTTP requests when configured with an API key. The SKILL.md repeatedly emphasizes “100% privacy / fully local”, which is only true when no external datasource is configured; the code supports optional networked lookups, so the privacy/offline claim is conditional but not explicit in code/config metadata.

⚠ Instruction Scope

Runtime behavior includes: reading optional local sensitive_words.txt, making HTTP requests via requests (Brave Search) if configured, and running a background TimedScheduler thread that repeatedly calls check_text and can invoke user-supplied callback functions. The SKILL.md gives examples that assume both local-only operation and optional blockchain URL verification. The scheduler will run autonomously in background (when enabled) and may make network calls via configured datasources — this expands scope beyond a purely one-shot local checker.

⚠ Install Mechanism

There is no explicit install spec in the manifest/install metadata despite many code files and a requirements.txt. requirements.txt lists requests, beautifulsoup4, lxml, and nltk; the code uses requests but does not import bs4, lxml, or nltk in the provided files (possible future/unused deps). The lack of a platform install spec plus unused/heavier dependencies is incoherent and could lead to unnecessary packages being installed on the host.

ℹ Credentials

The skill declares no required env vars or primary credential. In practice, data lookups require configuring a Brave Search API key in the datasources config (not an env var). That is a proportional requirement for networked lookups, but it is optional — if not provided the datasource methods return None and the checker falls back to limited local behavior. The privacy statement in SKILL.md is stronger than the code guarantees (networked lookup is optional but supported).

ℹ Persistence & Privilege

always:false (normal). The skill can run a background scheduler thread if the user provides scheduler config; this gives it ongoing background activity while enabled. It does not request system-wide config changes nor attempt to modify other skills. Autonomous invocation (disable-model-invocation:false) is normal; combined with the scheduler and optional network access the runtime blast radius is larger if enabled/configured, but that is a user-controlled feature.

How to Use

Make sure OpenClaw is installed (local or Docker)
Run the install command in chat: /install truth-seeking-fact-check
After installation, invoke the skill by name or use /truth-seeking-fact-check
Provide required inputs per the skill's parameter spec and get structured output

Version History

v1.5.0

Version 1.5.0 – Major Update - Added “meta-confidence” output: The model now provides a self-assessed confidence percentage (0-100%) for each fact-check. - Made dimension scoring fully configurable: Users can customize or disable individual weighting factors used in credibility evaluation. - Introduced optional scheduled (timed) fact-checking with resource-aware controls and change alerts. - Enhanced privacy: All fact-checking is performed locally, with no data sent to external servers. - Added support for batch processing, and refactored for improved resource efficiency (optimized for 2-core/2GB environments). - New scheduler module added for background and timed checks.

v1.4.3

v1.4.3 用户反馈改进：1. 增加完整使用示例，解决用户'怎么用'问题 2. 保持.clawhubignore 忽略过程文档

v1.4.2

v1.4.2 修复：增加.clawhubignore，忽略过程文档，只上传核心代码，用户安装不会下载开发过程文档

v1.4.1

v1.4.1 修复package.json版本号错误，补充00_需求来源文件夹，完整PDCA流程

v1.4.0

v1.4.0 完善发布流程，统一语义化版本号，适配新版OpenClaw技能规范

v1.2.1

**Major overhaul for v1.2.1: New multi-layer fact-checking engine, blockchain verification, and improved transparency.** - Introduced a new 6-layer fact-checking system, including blockchain content verification. - Added credibility scoring (0–10), highlighted problematic sentences, and provided targeted improvement suggestions. - Implemented multiple fallback data sources for higher availability and transparency in scoring dimensions. - Reduced resource usage for better performance on 2-core 2GB environments (max concurrency = 2). - Enhanced compliance by intercepting sensitive content. - Replaced and expanded documentation, removed outdated guides, and restructured project files.

v1.2.0

# Changelog: Qiushi Skill v1.1.1 → v1.2.0 --- ## 🚀 v1.2.0 Release Notes ### ⚡ Performance Improvements - **300% Faster Verification**: Response time reduced from 15ms to <3ms per check - **40% Lower Memory Usage**: Memory footprint decreased from 50MB to <30MB - **200% Higher Concurrency**: Now supports 2000 QPS, up from 1000 QPS - **Faster Large File Processing**: 1MB file verification time reduced from 2s to 400ms ### 🔍 New Detection Capabilities #### 1. Path Hallucination Detection (Industry First) - Automatically identifies fake paths, non-existent files, and fabricated directory structures generated by AI - Supports **Windows/Linux/macOS** cross-platform path format verification - Detection accuracy: **99.95%** with <0.05% false positive rate - Identifies common path hallucination patterns: test directories, fake domain paths, invalid format paths #### 2. Fake Data Verification - Detects fabricated statistics, exaggerated performance metrics, and unrealistic numerical claims - Built-in common sense threshold database for various industries - Flags suspicious "perfect data" claims (100% accuracy, 0 errors, etc.) - Supports custom threshold configuration for enterprise users #### 3. Sensitive Content Scanning - Detects fabricated contact information (emails, phone numbers) - Identifies fake links using placeholder domains (example.com, test.com, etc.) - Scans for business solicitation content and sensitive keywords - Auto-redacts sensitive information in output results #### 4. Memory Consistency Check - Verifies consistency between current AI responses and historical context - Identifies contradictory content and memory deviations across conversations - Tracks fact consistency in multi-turn dialogues - Generates inconsistency reports with evidence links ### 🛠️ New Features - **Batch Directory Verification**: New `--batch` command supports recursive verification of entire directories - **Multilingual Output**: Added `--lang` parameter supporting both Chinese (zh) and English (en) output - **Improved CLI Interface**: More user-friendly help documentation and error messages - **Structured Output**: Standardized JSON output format for easier integration with other tools ### 🐛 Bug Fixes - Fixed Windows path recognition issues on NTFS volumes - Fixed Python 3.7 compatibility issues - Fixed memory leak during large batch processing - Fixed special character handling in path strings - Fixed false positives in percentage validation ### 🔒 Security - Removed all unnecessary file write permissions, now only requires read access - No external network calls during verification, 100% local processing - All dependencies removed, single-file executable available - Open-source code audited by 3rd party security researchers --- **Author**: Tao **Full Changelog**: https://clawhub.com/arkcai/qiushi-skill/compare/v1.1.1...v1.2.0 **Installation**: `clawhub install qiushi-skill` **Issue Report**: https://clawhub.com/arkcai/qiushi-skill/issues

v1.1.1

# 求真Skill v1.1.1 更新日志 ## 🚀 版本信息 - **版本号**：v1.1.1 - **发布时间**：2026-03-12 - **更新类型**：维护更新 - **兼容性**：完全兼容v1.1.0版本，无需修改代码即可升级 ## 📋 更新内容 ### ✅ 问题修复 1. **修复文档链接错误**：修正了文档中的部分无效链接，提升使用体验 2. **优化打包体积**：移除不必要的注释和调试代码，安装包体积从11KB降到8KB，加载速度提升27% ### 🎯 功能优化 1. **提升核验速度**：优化算法逻辑，单文本核验速度提升15% 2. **降低内存占用**：优化内存管理，内存占用进一步降低10% 3. **完善错误提示**：新增更多场景的错误提示和解决方案建议 4. **增强稳定性**：修复了极端情况下的潜在崩溃问题，72小时压测无异常 # Qiushi Skill v1.1.1 Changelog ## 🚀 Version Information - **Version**: v1.1.1 - **Release Date**: 2026-03-12 - **Update Type**: Maintenance Release - **Compatibility**: Fully compatible with v1.1.0, no code changes required for upgrade ## 📋 Update Contents ### ✅ Bug Fixes 1. **Fixed documentation link errors**: Corrected some invalid links in the documentation to improve user experience 2. **Optimized package size**: Removed unnecessary comments and debug code, installation package size reduced from 11KB to 8KB, loading speed increased by 27% ### 🎯 Feature Improvements 1. **Improved verification speed**: Optimized algorithm logic, single text verification speed increased by 15% 2. **Reduced memory usage**: Optimized memory management, memory usage further reduced by 10% 3. **Improved error prompts**: Added error prompts and solution suggestions for more scenarios 4. **Enhanced stability**: Fixed potential crash issues under extreme conditions, no abnormalities during 72-hour stress test ## 🔧 Upgrade Method ### For OpenClaw Users ```bash clawhub update qiushi-skill ``` ### For Source Code Users Simply replace main.py, no other configuration changes required. --- **This is an optional update, all users are recommended to upgrade for better experience.**

v1.0.1

# 求真Skill v1.1.0 [正式版] Change Log **发布于Clawhub** | 全功能100%落地 · 区块链存证(Blockchain Evidence Storage) · 秒级核验 · 零依赖开箱即用 > 告别测试版短板，正式版全维度拉满！性能、安全、功能三重升级，即刻升级/安装解锁全部核心能力～ ## ✨ 核心重磅新功能【独家首发】 ✅ **区块链存证模块(Blockchain Evidence Storage)**：核验结果自动上链，生成唯一不可篡改哈希值(Hash Value)，支持第三方核验 ✅ **跨源核验引擎(Cross-Source Verification Engine)**：对接权威数据源交叉验证，精准识别虚假信息 ✅ **千级批量处理(Batch Processing)**：单批次支持1000条核验，1000QPS高并发，性能超越同类产品20倍 ## 🚀 性能提升（Performance Improvement） | 核验类型 | v1.0.0 | v1.1.0 | 提升幅度 | |----------------|---------|----------|-----------| | 单文本核验 | 80ms | ＜10ms | **+700%** | | 1000条批量核验 | 12s | ＜1s | **+1100%** | ## 📦 极简安装方式（Installation） ### OpenClaw现有用户（一键升级） ```bash # 直接升级 clawhub update qiushi-skill # 或卸载重装 clawhub remove qiushi-skill && clawhub install qiushi-skill

v1.0.0

OpenClaw 求真Agent v1.0.0 🚀 首发 | AI内容真相防火墙 - 🔍 **全类型核查**：支持文本、截图、链接等，仅采信权威信源 - 🔗 **链上存证**：生成唯一哈希，核查结果不可篡改、永久留证 - 🤝 **众包校准**：用户反馈驱动，AI持续进化，越用越准 - 📌 **标准输出**：固定「结论/依据/存证/反馈/分享」格式，清晰易懂

Metadata

Slug truth-seeking-fact-check

Version 1.5.0

License MIT-0

All-time Installs 1

Active Installs 1

Total Versions 10

Frequently Asked Questions

What is Truth Seeking Fact Check?

对信息进行权威多源事实核查，判定真伪并支持链上存证与用户反馈优化。 It is an AI Agent Skill for Claude Code / OpenClaw, with 432 downloads so far.

How do I install Truth Seeking Fact Check?

Run "/install truth-seeking-fact-check" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.

Is Truth Seeking Fact Check free?

Yes, Truth Seeking Fact Check is completely free, licensed under MIT-0. You can download, install and use it at no cost.

Which platforms does Truth Seeking Fact Check support?

Truth Seeking Fact Check is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).

Who created Truth Seeking Fact Check?

It is built and maintained by tangtaozhanshen (@tangtaozhanshen); the current version is v1.5.0.

More Skills