How to Configure New Agents
This page describes the current OpenSIN path for introducing new agents without creating duplicate repos or drifting away from canonical ownership.
IMPORTANT
Before creating any new agent surface, read OpenSIN-overview/docs/CANONICAL-REPOS.md. Many new agent ideas belong inside an existing Team-SIN-* monorepo instead of a brand-new repository.
Step 1: Decide Where the Agent Belongs
Use this rule first:
- if the agent belongs to an existing team, add it to that
Team-SIN-*repo - if it requires a new standalone agent repo, start from
Template-SIN-Agent - if you are unsure, open a repo proposal in
OpenSIN-overviewbefore coding
Step 2: Use the Canonical Template
OpenSIN standardizes new standalone agents from:
This template is the canonical blueprint for new A2A-capable agents.
Step 3: Register the Agent in Canonical Config
Agent registration and model routing belong in the canonical OpenCode stack:
OpenSIN-AI/Infra-SIN-OpenCode-Stack- local
~/.config/opencode/*files when working on a machine-specific setup
Update only the canonical config locations. Do not invent parallel config repos.
Step 4: Define the Agent Contract
Every production-grade agent needs at minimum:
agent.json.well-known/agent-card.jsonAGENTS.md- governance metadata required by the template
Step 5: Verify Ownership and Exposure
Before claiming the agent is ready:
- confirm the owning repo is canonical
- confirm the docs point to the right repo and surface
- confirm any public endpoint is verified before publishing it in docs
Step 6: Publish Safely
Public docs must not expose internal IPs, internal dashboards, or unverified production endpoints. If an endpoint is not publicly verified, document it as internal/unverified instead of presenting it as final truth.