Running local agents

The C-Node is intentionally a thin OpenAI-shaped HTTP layer over Morpheus. That makes it a clean target for any agent framework — LangChain, LlamaIndex, custom tools, OpenClaw, etc.

Per-agent users

Don’t share the admin user with agents. Add a per-agent user with a method whitelist:

curl -X POST 'http://127.0.0.1:8082/auth/users' \
  -H 'Authorization: Basic <base64(admin:pw)>' \
  -H 'Content-Type: application/json' \
  -d '{
    "username": "research-agent",
    "password": "rotate-me",
    "methods": ["chat", "open_session", "close_session", "get_balance"]
  }'

This writes a new rpcauth= and rpcwhitelist= line to proxy.conf. Full method list and semantics: API auth.

Session policies

Pattern	When to use	Notes
Pre-open one long session per agent run	Repeated short prompts	Lower per-prompt latency; one cleanup at end
Open per task	Bursty, infrequent tasks	Higher gas overhead, simpler state
Pool of pre-opened sessions	High-throughput multi-agent	Fan agents across sessions; close idle ones

Always close sessions explicitly when work is done. Unused stake returns in the close txn; the used stipend day-locks in userStakesOnHold until the next UTC day (claim via withdrawUserStakes). Plan agent float for gross daily stake — used MOR is not reusable the same UTC day. See Sessions: stake, close, claim.

Streaming vs non-streaming

Agents that need full deterministic JSON should use stream: false. Tool-calling chains that benefit from token streaming (e.g. user-facing typing UX) should set stream: true. The proxy-router supports both transparently — see Chat.

Rate limiting

The proxy-router does not currently expose per-user rate limits. If you need hard limits, terminate agent traffic at a reverse proxy (nginx limit_req, Caddy rate limiter, etc.) in front of :8082.

Observability

Tag each agent with its own User-Agent header for log filtering.
Run a small log shipper to capture ./data/proxy-router.log lines.
Use GET /v1/models/attestation if your agents target tee-tagged models — surface attestation state in your dashboards.

Failure recovery

Symptom	Recovery
401 / 403 from C-Node	Re-auth or rotate password
429 from C-Node	Back off; check provider/model concurrency
Empty response stream	Provider may be down — close session, retry on a different model
`session not found`	Session expired or closed; open a new one

For a more concrete worked example, see Gateway for Everclaw.

Get started

Inference API (hosted)

Concepts

Consumers

Prosumers

Providers - Resale

Providers - Full P-Node

AI knowledge

Ecosystem

Reference

Per-agent users

Session policies

Streaming vs non-streaming

Rate limiting

Observability

Failure recovery

​Per-agent users

​Session policies

​Streaming vs non-streaming

​Rate limiting

​Observability

​Failure recovery

Per-agent users

Session policies

Streaming vs non-streaming

Rate limiting

Observability

Failure recovery