eval-harness

6 main sessions · 0 sub-agent runs

LLM evaluation harness — a RAG faithfulness metric with evidence, a parallel resumable runner, HTML run reports, and eval gates in CI.

pythonllmevaluationragasyncioci

https://example.com/eval-harness ↗

Activity · last 365 days · eval-harness

Tool calls · eval-harness aggregate

Token usage · timeline548.8K total · 67% cache hit · warming up

Main sessions (6)

rag-faithfulness-metric — 2026-09-04

2026-09-04 · claude-opus-4-7

1 messages · 3 tool calls

updated 2 months ago

golden-dataset-curation — 2026-09-08

2026-09-08 · gemini-2.5-pro

1 messages · 2 tool calls

updated 8 weeks ago

parallel-runner-with-retries — 2026-10-02

2026-10-02 · qwen3-coder

1 messages · 3 tool calls

updated 4 weeks ago

html-report-for-eval-runs — 2026-10-30

2026-10-30 · gpt-5-codex

1 messages · 3 tool calls

updated 5 days ago

cost-tracking-per-run — 2026-11-02

2026-11-02 · qwen3-coder

1 messages · 3 tool calls

updated 2 days ago

run-evals-in-ci-on-prompt-changes — 2026-11-04

2026-11-04 · claude-sonnet-4-6

1 messages · 3 tool calls

updated yesterday