Overview
Make misuse harder than correct use.
BaseCred is a read-only reputation primitive. It aggregates signals from trusted on-chain and off-chain providers (Ethos, Neynar, Talent Protocol) and exposes them as neutral, deterministic context — not judgments.
BaseCred answers one question only:
“What signals are observable about this identity right now?”
It does not decide who to trust, who to block, or what action to take.
Core Principles
- Context, not decisions: BaseCred provides descriptive signals. Applications own all interpretation, thresholds, and enforcement.
- No judgment: We never label users, rank them, or produce final verdicts.
- Deterministic: The same input always results in the same output.
- Explicit absence: Missing data is surfaced clearly — never inferred or hidden.
- Transparent by design: Every field is traceable to a source. No hidden weighting. No black boxes.
What BaseCred Is Not
BaseCred is not an authorization engine, moderation system, trust oracle, or recommendation system.
If your product requires a final score or automatic action, BaseCred is intentionally the wrong tool.
Integration
You can integrate BaseCred via the SDK or public API. The response schema is versioned and stable (Semantic Versioning). See the Integration Guide for details.
See Context vs Decision to understand where BaseCred stops and your application begins.
If you are building an AI agent (e.g. OpenClaw), see the OpenClaw Integration for a quick-start guide and the Agent API Reference for full endpoint docs. Agents can self-register, verify via tweet, and check their owner's reputation autonomously.
If you are integrating a ZK agent flow, see the ZK Agent Integration page for policy hashes and proof payload examples.
Visit zkbasecred.xyz/human to see live data.