---
name: api.veritylayer.dev
description: VerityLayer exposes a single fact-checking endpoint that accepts a textual claim plus optional context and returns a structured verdict (supported or refuted), a confidence score, reasoning, a list of evidence sources, and a cryptographically signed receipt. It is designed for agents that need machine-verifiable, auditable fact-checking against established knowledge bases.
host: api.veritylayer.dev
---

# api.veritylayer.dev

api.veritylayer.dev is a fact-verification API host serving agents and pipelines that must validate discrete factual claims against authoritative sources. Its distinguishing feature is the signed cryptographic receipt attached to every verdict, enabling downstream systems to prove a claim was checked without re-running the verification. It does not perform real-time data lookups, opinion analysis, or continuous monitoring.

## When to use this host

Use this host when an agent needs to validate a specific, discrete factual claim and requires a structured, auditable result — particularly when a signed receipt is needed for downstream trust or compliance purposes. Do not use it for real-time or rapidly changing information such as live prices, current events, or breaking news; route those queries to live data APIs or news feeds instead. Do not use it for subjective judgments, opinion validation, or sentiment analysis. Because the host exposes only one logical skill (claim verification), it is not suitable as a general-purpose research or retrieval tool; pair it with a retrieval or search host when broader information gathering is needed before a specific claim can be formulated.

## Capabilities

### Factual Claim Verification

Accepts a claim and optional context, evaluates it against established knowledge sources, and returns a structured verdict with confidence score, reasoning, evidence list, and a tamper-evident signed receipt.

- **`verify-factual-claim`** — Verifies a factual claim against multiple sources and returns a supported/refuted verdict with confidence score, reasoning, evidence list, and a signed cryptographic receipt.

## Skill reference

### `verify-factual-claim`

**VerityLayer Quick Verify** — Submits a factual claim (and optional context) to VerityLayer and returns a supported/refuted verdict with confidence score, reasoning, evidence list, and a signed receipt.

*Use when:* Use when an agent needs to fact-check a specific factual claim against established knowledge and requires a machine-verifiable verdict with confidence score, reasoning, and a cryptographically signed receipt for auditability.

*Not for:* Do not use for opinion validation, subjective judgments, or real-time data lookups (e.g. current prices or live events); this endpoint verifies factual claims against established knowledge only.

**Inputs:**

- `claim` (string, required) — The factual claim to verify. Must be between 1 and 2000 characters. Oversized or blank input returns a free 422 error rather than a paid non-answer.
- `context` (string) — Optional additional context to accompany the claim, up to 2000 characters. May include background or framing to help the verifier assess the claim accurately.

**Returns:** Returns verdict ('supported'/'refuted'), a confidence score, a reasoning string, an array of evidence strings, and a signed receipt object with SHA-256 hashes of the claim, context, and output plus an ed25519 signature.

**Example:** `{"claim": "The Eiffel Tower is in Paris.", "context": "Please verify this claim and provide a detailed analysis that explains where the Eiffel Tower is located and any relevant geographic or administrative details."}`

---

### `verify-factual-claim`

**VerityLayer Claim Verifier** — Verifies a factual claim against multiple sources and returns a supported/refuted verdict with confidence score, reasoning, evidence list, and a signed cryptographic receipt.

*Use when:* Use when an agent needs to fact-check a specific claim against authoritative sources and requires a structured verdict with traceable evidence and a tamper-evident signed receipt for auditability.

*Not for:* Do not use for real-time or rapidly changing data (e.g., live prices, breaking news); this is a single-shot claim verification, not a streaming or continuous monitoring service.

**Inputs:**

- `claim` (string, required) — The factual claim to verify. Must be between 1 and 2000 characters.
- `context` (string) — Optional additional context to guide verification, up to 2000 characters. Null if not provided.

**Returns:** Returns a verdict ('supported'/'refuted'), a confidence score (0.99), a reasoning narrative citing sources, an evidence array with specific source attributions, and a signed receipt with SHA-256 hashes of the claim, context, and output for auditability.

**Example:** `{"claim": "The Eiffel Tower is located in Paris, France.", "context": "Verify the truth of this claim. Provide evidence-based reasoning, and include any relevant geographic/context details (e.g., administrative location, country) that would confirm or refute it."}`

---