examples-openclaw — OpenClaw + Ainfera

For OpenClaw builders. Drop-in NemoClaw-compatible inference backend with signed identity, drain-proof Wallet, and hash-chained audit.

Two-line change

nemoclaw config inference.backend=ainfera
nemoclaw config inference.api_key=ai_live_...        # https://app.ainfera.ai/signup

Your OpenClaw skills, tools, and orchestration loop keep working. You now have:

• Signed Agent Card identity (L1)
• Drain-proof per-call Wallet — survives prompt injection (L3)
• Hash-chained audit per call, verifiable offline (L4)
• 5 frontier models via one key: Claude Opus 4.7 · GPT-5.5 · Gemini 3.1 Pro · Grok 4 · Mistral Large 3

EU AI Act Annex IV ready — every call hash-chained, signed, exportable.

Quickstart

If you have OpenClaw / nemoclaw installed:

git clone https://github.com/ainfera-ai/examples-openclaw
cd examples-openclaw
export AINFERA_API_KEY=ai_live_...
./main.sh

Or with curl only — no OpenClaw install needed:

AINFERA_API_KEY=ai_live_... ./curl-example.sh

What this proves

OpenClaw's inference layer accepts Ainfera as a NemoClaw-compatible backend. Drop in the two config lines above and every skill's inference call now produces a signed AgentCard + Wallet receipt + AuditChain entry. No prompt or skill changes required.

See the OpenClaw Pool landing page for the full pitch + companion concept pages.

Verify the audit chain

After running, each call returns a receipt URL. You can also verify offline:

pip install ainfera-verify
ainfera-verify chain <your-agent-id>

License

Apache 2.0. See LICENSE.