- Home
- Skills
- Research & Analysis
- Multi-Source Credibility & Fact-Checker by Saki-AI
Multi-Source Credibility & Fact-Checker by Saki-AI
Verify claims with a multi-source research workflow, heuristic domain scores and English/Chinese emotional-language checks. Includes a local Node.js scorer. Free from the creator of Saki-AI and Saki-Panel.
Free
Works with the AI Agents you already use
Multi-Source Credibility & Fact-Checker by Saki-AI
Example session with this skill installed
Run the bundled credibility scorer demo and explain all five metrics. Treat the supplied URLs and stances as a synthetic fixture, not as verified live evidence.
- Read your context and instructions
- Compiled the multi-source credibility
Synthetic scoring demo
Command: node scripts/credibility-scorer.cjs --demo
The bundled fixture contains a sensational claim and three preassigned source stances. No web pages are fetched by this command. The URL strings are scoring inputs and must not be treated as retrieved citations.
| Output | Value |
|---|---|
| Score | 21 / 100 |
| Verdict | likely_false |
| Evidence | 10 |
| Source authority | 74 |
| Diversity | 76 |
| Consistency | 86 |
| Emotion score | 14 |
The low evidence score reflects the fixture stances, which lean against the claim. Domain authority uses a fixed heuristic. Diversity rewards distinct domains and search engines. Consistency measures alignment of weighted stances. The emotion component penalizes sensational language; it is not evidence that the claim is false.
For a real investigation
- Normalize the claim and define what would verify or contradict it.
- Retrieve independent primary sources with your agent web tools.
- Check publication dates, relevant passages and source independence.
- Assign stances based on the actual content, then run the scorer.
- Report verified citations and unresolved uncertainty alongside the heuristic result.
A numerical score supports review; it does not replace source verification.
Connects securely to your tools. The creator never sees your data.
About this skill
Turn web claims into structured evidence reviews
A free fact-checking workflow and deterministic scoring helper extracted from Saki-AI. Separate a falsifiable claim from emotional framing, search independent sources, assign evidence stances and produce a transparent report.
Included
- SKILL.md instructions for claim normalization, cross-source searches and evidence synthesis.
scripts/credibility-scorer.cjs: heuristic domain authority, English/Chinese emotional-language signals, source diversity and weighted evidence scoring.- JSON output with a score, verdict, component metrics and per-source values.
- Apache-2.0 license; no npm packages required.
How it works
An agent uses its own web tools to retrieve and assess sources, then passes source URLs, stances and engine names to the scorer. The local script does not fetch pages or independently establish that a claim is true. Scores are transparent heuristics, not calibrated probabilities.
Start with node scripts/credibility-scorer.cjs --demo for a synthetic fixture, then supply your own evidence JSON for real claims.
Explore both projects
This free skill is shared by the creator of Saki-AI and Saki-Panel. Visit both repositories, try the projects, and consider giving both a GitHub star to support continued development.
- Saki-AI: https://github.com/EthanChan050430/Saki-AI
- Saki-Panel: https://github.com/EthanChan050430/Saki-Panel
Find more skills like this on the Claude Skills Marketplace, the Codex Skills Marketplace and the Gemini Skills Marketplace, or browse Claude Skills for Research & Analysis.
Scores are heuristics, not probabilities or proof. The script does not fetch pages or verify supplied stances. Source independence, actual content and publication dates must be checked by the agent. Built-in demo URLs are fixture inputs, not verified citations.
How to use
- 1
Connect your agent to the Agensi MCP
Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup
- 2
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 3
Ask for what you need
Your agent loads this skill and uses it on your task. It always asks before buying a skill.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill goes through the trust checks below.
Security scanned
Checked by our 8-point scan before listing
Fresh listing
Recently published to Agensi
Free to use
No account required to browse
Trust & safety
Security scan
Security scan passed 15 days ago
- Free with an Agensi account
- Software
Node.js 18+
Agent web searchorbrowser tool
Permissions
The scorer itself runs locally without network access. Real claim verification uses agent web tools to retrieve sources. No paid API or account is required by the scorer.