Skip to content

v1.3.0 — Fireworks AI Provider

Latest
Latest

Choose a tag to compare

View all tags
@namanvats-dev namanvats-dev released this 16 Jun 06:45
v1.3.0
1d7dc3e
This commit was created on GitHub.com and signed with GitHub’s verified signature.
GPG key ID: B5690EEEBB952194
Verified
Learn about vigilant mode.

✨ What's New

Fireworks AI is now a first-class provider — usable both as an evolution harness backend and as the LLM scorer (#50).

  • Authentication: set FIREWORKS_API_KEY. EvoSkill auto-mirrors it to FIREWORKS_AI_API_KEY for litellm-based harnesses, so you only set one key.
  • Automatic provider detection from the model string — fireworks/, fireworks-ai/, fireworks_ai/ prefixes, or a bare Fireworks model id (accounts/fireworks/models/...) all resolve to Fireworks.
  • Provider-alias handling so each consumer gets the id it expects: fireworks-ai (OpenCode / models.dev), fireworks_ai (OpenHands / litellm), fireworks (LLM scorer). All point at the OpenAI-compatible endpoint https://api.fireworks.ai/inference/v1.
  • Forwarded to Docker and Daytona runs, so remote/containerized evolution picks up the Fireworks key automatically.

🚀 Usage

Set your key:

export FIREWORKS_API_KEY=your-key-here

📝 Notes

  • Goose has no built-in Fireworks provider — use a manual OpenAI-compatible configuration.
  • New tests cover provider-key resolution and model-string normalization (tests/test_provider_auth.py, tests/test_cli_shared.py).

🔄 Compatibility
No breaking changes — purely additive. Existing Anthropic / OpenAI / OpenRouter / Google setups are unaffected.
Full changelog: v1.2.2...v1.3.0

Assets 2
Loading