The Inference of AI Agents. Drain-proof wallet for AI agents.
Your agent's wallet survives prompt injection. One Agent Card across providers. Every call cryptographically audited.
pip install ainfera| Repo | What |
|---|---|
| api | Platform service — L1–L5 — api.ainfera.ai |
| sdk | Python + TypeScript SDK — pip install ainfera |
| mcp-server | MCP tool surface — mcp.ainfera.ai |
| verify | Offline AuditChain verifier — pip install ainfera-verify |
| specs | Open specs (CC-BY 4.0) — Ontology, ATS, AAMC, Annex IV, A2A |
| ainfera-os | Internal dogfood — 5 agents calling the public SDK |
| web | Marketing site source — ainfera.ai |
| Pool | Repo | What |
|---|---|---|
| Hermes | examples-hermes | hermes-agent + Ainfera in 2 env vars |
| OpenClaw | examples-openclaw | OpenClaw / NemoClaw inference backend in 2 config lines |
| OpenAI-compat | examples-openai-compat | Any OpenAI SDK · 2-env-var swap |
| LangChain | examples-langchain | LangChain agent · drop-in chat model |
| CrewAI | examples-crewai | CrewAI flow · per-agent signed identity |
| Google ADK | examples-google-adk | Google Agent Developer Kit · auditable inference |
| MCP | examples-mcp | MCP client → mcp.ainfera.ai tools |
JWS RFC 7515 · did:web · x402 Foundation (LF) · NIST AI Agent Standards · EU AI Act Annex IV · Mastercard Verifiable Intent · Open Agent Identity (OAI)
Code: Apache 2.0 · Specs: CC-BY 4.0