Skip to content
repository radar
Agents & Orchestration

🧪EvalView

ACTIVE STEADY

hidai25/eval-view · homepage ↗

Regression testing for AI agents. Snapshot behavior,diff tool calls,catch regressions in CI. Works with LangGraph, CrewAI, OpenAI, Anthropic.

⭐ Emerging: 128 stars, gaining about 2 a week

View on GitHub ↗

repo profile

vintage 2025 1 year old
delivery product
language Python
license Apache-2.0

momentum

total stars 128
stars added last week +2
weekly downloads 354 /wk · PyPI
commits / week 1 cooling · -75% vs prior mo
issues closed 99% ~6d to close, median
contributors 15
release cadence fortnightly
last activity 20d ago

durability

backing community / independent
openness permissive
bus factor 1 solo
top-author share 67% 6 mo

bus factor = how many people it takes to cover more than half the commits (6 months). 1 is a solo project; higher means the work is spread across a team. top-author share is the single busiest author's slice of those commits.

since we covered it

monthly average + 7/mo (+6%/mo) · + 14 total since PR#36

why it's a big deal

  • Gives teams a merge-time regression gate for AI agents, catching behavior drift in CI before it ships rather than after users hit it.
  • Requires zero assertions written upfront; it snapshots what the agent currently does and flags any change, which lowers the cost of adding coverage to existing agents.
  • Framework-agnostic, working with LangGraph, CrewAI, OpenAI, Claude, Mistral, Ollama, MCP, or any HTTP API endpoint.

under the hood

  • Written in Python and usable as both a CLI and an importable library, driven by a two-command workflow of evalview snapshot to record a baseline and evalview check to compare against it.
  • Diffs the full tool-call trajectory, the tools called, their parameters, and execution order, with deterministic comparison that runs offline without an API key and an optional LLM judge for output-quality scoring.
  • Ships a GitHub Actions integration that posts PR comments with diffs plus cost and latency deltas, and supports multi-variant baselines of up to 5 valid paths to handle non-determinism.

Radar summary, generated from the project's public sources

star history

PR#36 · 114128 now Nov 2025Aug 2026
  1. PR#36 114 2026-06-17
  2. now 128 + 14 since first covered

curve is sampled from GitHub's star history, plus our own daily readings since we covered it; the dashed stretch is before we first covered it, the solid line since. figures at coverage are the numbers we printed then (approx.), current count is live.

starbound

On our Starbound watchlist: shipping hard from a cold start, before its stars have caught up.

+48 starbound score output 61 · clout 22
Aug 2025 Jul 2026
  • output, commits & releases
  • clout, star velocity

output = commits & releases; clout = star velocity, both 0 to 100 monthly indices. Shipping this far ahead of its stars from a cold start is our leading indicator. The Starbound watchlist →

covered in

  • PR#36 2026-06-17 below the radar

    A behavior regression gate for agents, the Playwright for AI agents

similar projects

compare these →
  • 📚 Haystack

    205× the stars

    Open-source AI orchestration framework for building context-engineered, production-ready LLM applications. Design modular pipelines and agent workflows with explicit control over retrieval, routing, memory, and generation. Built for scalable agents, RAG, multimodal applications, semantic search, and conversational systems.

    26k ACTIVE
  • 🔗 Composio

    TypeScript · 232× the stars

    Composio powers 1000+ toolkits, tool search, context management, authentication, and a sandboxed workbench to help you build AI agents that turn intent into action.

    30k ACTIVE
  • 🦌 DeerFlow

    626× the stars

    An open-source long-horizon SuperAgent harness that researches, codes, and creates. With the help of sandboxes, memories, tools, skill, subagents and message gateway, it handles different levels of tasks that could take minutes to hours.

    80k ACTIVE

comments

Sign in with GitHub to add your blip on EvalView.

loading comments…