name: Live Agent Evaluation
on:
workflow_dispatch:
inputs:
model:
description: Anthropic model used by the paid smoke test
required: false
default: claude-haiku-4-5-20251001
permissions:
contents: read
jobs:
live-anthropic:
name: Live Anthropic smoke
runs-on: ubuntu-latest
env:
ANTHROPIC_API_KEY: ${{ secrets.ANTHROPIC_API_KEY }}
CLAUX_EVAL_MODEL: ${{ inputs.model }}
steps:
- uses: actions/checkout@v4
- name: Require evaluation secret
shell: bash
run: |
if [ -z "$ANTHROPIC_API_KEY" ]; then
echo "Repository secret ANTHROPIC_API_KEY is required" >&2
exit 1
fi
- name: Install Rust
uses: dtolnay/rust-toolchain@stable
- name: Run paid live provider smoke
run: cargo test evals::live_anthropic_smoke -- --ignored --nocapture