LLM & Agents Quickstart

This section is for teams building LLM or agent tooling around Feral File APIs.

It is a support layer, not a separate product track.

• What this is: a thin setup guide for tool-calling with Feral File APIs.
• Why use it: connect an agent safely without duplicating protocol or CLI logic.
• What to do next: choose FF1 CLI or DP-1 as your primary execution path.

What this is for

• Connect an LLM tool to OpenAPI endpoints.
• Keep calls deterministic and small.
• Route execution to the right primary docs path.

Use these primary paths

• Playlist generation and playback workflows: FF1 CLI: Start Here
• Protocol correctness, schemas, validator, and feeds: DP-1: Start Here

Minimal setup

1. Import OpenAPI schema:
2. https://feralfile.com/.well-known/openapi.json
3. Add system instructions that keep requests scoped and explicit.
4. Keep request limits small to avoid oversized context and unstable outputs.
5. Validate generated playlist payloads with DP-1 Validator Quickstart.

If your agent should run FF1 CLI directly, use this ready prompt as your starting point:

• https://github.com/feral-file/ff1-cli/blob/main/skills/ff1-control/SKILL.md

This keeps the flow simple: status -> config validate -> build -> validate -> send/publish, and reports failing command + code when a step breaks.

Guardrails

• Do not treat this section as a replacement for CLI docs.
• Do not duplicate DP-1 schema or feed operation docs in agent prompts.
• Keep protocol operations in DP-1 pages and command workflows in FF1 CLI pages.
• Do not claim end-to-end DP-1 v1.1.0 parity across all tools unless verified in upstream repos.

Next step

Choose your execution path: FF1 CLI first run or DP-1 integrator start.