Using Model Endpoint
Apps & Automationusing-model-endpointUsing Model Endpoint is an Agent Skill for apps & automation published in the companion-inc/feynman repository on GitHub: a SKILL.md instruction file an AI agent can load to do this task.
Call a configured Feynman model endpoint and interpret its response. Use when a task needs inference from a registered endpoint, remote model API, local model service, or custom connector-backed predictor.
How to use this skill
Bring this guide into your coding agent with a prompt tailored to the tool you use.
- Open your project in Codex.
- Copy and paste the prompt below.
- Review files and risks before approving any action.
Inspect the Agent Skill at https://github.com/companion-inc/feynman/blob/HEAD/skills/using-model-endpoint/SKILL.md and all supporting files. Report provenance, permissions, scripts, dependencies, and security risks. Propose an exact installation into .agents/skills/using-model-endpoint, but wait for my approval before writing files or running scripts.
Copying this prompt does not install or run the skill. Review third-party files before use.
Using Model Endpoint
Use this skill after a model endpoint is already configured or explicitly chosen.
Workflow:
- Confirm endpoint identity, auth status, input schema, output schema, model version, and rate/size limits.
- Build a minimal request with explicit inputs and no hidden context.
- Save request metadata, response, latency, status, and parsing code as artifacts without exposing secrets.
- Validate response shape and handle model errors as evidence, not as missing work to hide.
- Interpret predictions separately from source-backed facts.
When the endpoint is not configured, switch to managed-model-endpoints or compute-env-setup instead of pretending inference ran.