Claim Verification
A claim + (optional) sources → a verdict with evidence and a calibrated
confidence.
Purpose
Before a claim ships in a post, brief, or video, it should be checked. This skill
takes a single claim, gathers or accepts sources, and returns a structured verdict
that separates what the evidence supports from what it doesn't — with honest
confidence, not false certainty.
When it fires
- Keyword triggers:
verify, fact-check, is this true, confirm,
source check, debunk
- File triggers:
claims.md, fact-check/**
- Command triggers:
/claim-verification
Inputs
| Input |
Required |
Notes |
claim |
yes |
The exact statement to check. |
sources |
no |
Provided URLs/quotes; else the skill gathers them. |
strictness |
no |
standard (default) or high (primary sources only). |
Workflow
- Restate the claim precisely — pin the testable assertion; flag ambiguity.
✅ Check: the claim is falsifiable as stated.
- Gather evidence — find/accept sources; prefer primary and authoritative.
- Weigh — does the evidence support, refute, or only partly cover the claim?
Note source quality and any conflict.
- Verdict + confidence — assign one of the four verdicts and a confidence
level with a one-line rationale.
- Cite — quote the decisive evidence with locations.
Output contract
{
"claim": "string — restated precisely",
"verdict": "supported | refuted | partially-supported | unverifiable",
"confidence": "low | medium | high",
"evidence": [{ "quote": "string", "source": "string" }],
"rationale": "string",
"caveats": ["string"]
}
Tools & MCP
- Tools:
WebSearch, WebFetch, Read.
- MCP dependencies: optional docs/library MCP for technical claims.
Quality bar
Example
Input: "The James Webb Space Telescope orbits the Earth." → verdict refuted
(it orbits the Sun–Earth L2 point), high confidence, with a cited NASA source.
Built on SIP — Starlight Intelligence Protocol
Substrate: starlightintelligence.org/protocol v1.1.0
Layers used: [file-contract, attestation, sovereignty]
Vertical: starlight-agent-skills · portable capability layer
1---2name: claim-verification3description: Fact-check a claim against sources and return a verdict with evidence and confidence — supported, refuted, partially-supported, or unverifiable. Use when asked to 'verify', 'fact-check', 'is this true', or before publishing a statistic or scientific claim.4---56# Claim Verification78> A claim + (optional) sources → a verdict with evidence and a calibrated9> confidence.1011## Purpose1213Before a claim ships in a post, brief, or video, it should be checked. This skill14takes a single claim, gathers or accepts sources, and returns a structured verdict15that separates what the evidence supports from what it doesn't — with honest16confidence, not false certainty.1718## When it fires1920- Keyword triggers: `verify`, `fact-check`, `is this true`, `confirm`,21 `source check`, `debunk`22- File triggers: `claims.md`, `fact-check/**`23- Command triggers: `/claim-verification`2425## Inputs2627| Input | Required | Notes |28|-------|----------|-------|29| `claim` | yes | The exact statement to check. |30| `sources` | no | Provided URLs/quotes; else the skill gathers them. |31| `strictness` | no | `standard` (default) or `high` (primary sources only). |3233## Workflow34351. **Restate the claim precisely** — pin the testable assertion; flag ambiguity.36 ✅ Check: the claim is falsifiable as stated.372. **Gather evidence** — find/accept sources; prefer primary and authoritative.383. **Weigh** — does the evidence support, refute, or only partly cover the claim?39 Note source quality and any conflict.404. **Verdict + confidence** — assign one of the four verdicts and a confidence41 level with a one-line rationale.425. **Cite** — quote the decisive evidence with locations.4344## Output contract4546```json47{48 "claim": "string — restated precisely",49 "verdict": "supported | refuted | partially-supported | unverifiable",50 "confidence": "low | medium | high",51 "evidence": [{ "quote": "string", "source": "string" }],52 "rationale": "string",53 "caveats": ["string"]54}55```5657## Tools & MCP5859- Tools: `WebSearch`, `WebFetch`, `Read`.60- MCP dependencies: optional docs/library MCP for technical claims.6162## Quality bar6364- [ ] Verdict matches the evidence; `unverifiable` is used honestly.65- [ ] Confidence is calibrated, not defaulted to "high".66- [ ] Every verdict-driving quote is real and located.67- [ ] Source quality is considered, not just source count.6869## Example7071Input: "The James Webb Space Telescope orbits the Earth." → verdict `refuted`72(it orbits the Sun–Earth L2 point), high confidence, with a cited NASA source.7374---7576Built on SIP — Starlight Intelligence Protocol77Substrate: starlightintelligence.org/protocol v1.1.078Layers used: [file-contract, attestation, sovereignty]79Vertical: starlight-agent-skills · portable capability layer