AI Testing & Debugging

Compare community-submitted benchmarks for local AI models

The LLM Evaluation Framework

Observal is an Observability and Evaluation platform for human-in-the-loop agents

See what your AI agents are doing behind the scenes

Independent audits for AI agent behavior and alignment

Main repository for Datadog Agent

The fastest path to AI-powered full stack observability, even for lean teams

Better Prometheus alerts for Kubernetes - smart grouping, AI enrichment, and automatic remediation

Agent analytics platform for monitoring AI agent sessions, performance, knowledge gaps, corrections, resolution, and business impact

Local code review studio for AI agent-generated changes with line comments and MCP handoff

Autonomous web app testing agent that explores flows, finds bugs, and generates self-healing tests from a URL

LLM engineering platform for gateway routing, observability, evaluations, and prompt optimization

Mac tool that turns screenshots into JSON readable by CLI agents

Automated bug tracking, error monitoring, and contextual reporting platform for modern web applications

Expect lets agents test code in a real browser by scanning diffs, generating plans, and running execution workflows from one command

TestSprite is an AI testing agent that can plan, write, execute, debug, and report software tests end to end

Ogoron is an automated testing platform where autonomous agents plan, generate, and maintain tests directly from your codebase

QA.tech uses AI agents to explore web apps, generate complete test suites, and run QA checks on release, schedule, or manual triggers

End-to-end testing and observability for conversational AI. Run pre-production simulations and monitor production conversations for voice and chat agents

Glassbrain helps teams visually debug LLM apps by capturing every OpenAI, Anthropic, and LangChain call and replaying failed runs

Future AGI helps teams build, evaluate, optimize, and monitor LLM and AI agent applications with multimodal quality testing and observability

Coding-agent observability for session costs, tokens, cache usage, tool calls, file edits, and terminal replays across Claude Code, Codex, and more

Waydev is an engineering analytics platform for measuring delivery performance, DORA metrics, and productivity trends

Simple, open-source LLM tracing for AI agents. Track prompts, completions, latency, token usage, and cost

Render AI chat data and Codex logs into filterable browser timelines

Regression testing and semantic diff platform for agent apps to catch behavior changes before release

Open-source debugging and run-inspection tool for AI agent workflows, prompts, tool calls, and runtime state

Open-source agent engineering platform with traces, evals, and metrics for debugging AI agents

AI-assisted release management for release notes, launch checklists, and rollout communication

AI agent that reproduces GitHub issues and returns evidence-backed reports

Debug AI agents by recording, replaying, forking, and sharing runs

Open-source AI agent monitoring and quality platform that detects failure modes and helps developers fix agent issues before production

Dedicated QraftAI agent for generating and running cross-browser QA tests

Product-drift review for AI-written pull requests

Real-time production evaluation for AI agent runs

OpenTelemetry-based tracing and observability for prompts, model calls, tool spans, tokens, latency, and errors in production AI applications

AI agent that monitors production logs, detects incidents, and opens reviewable fix PRs

AI tool that reviews pull requests, analyzes CI failures, and submits fixes

Runs pull requests in ephemeral environments and reviews them with runtime evidence

Lets coding agents and CI publish screenshots, recordings, and ready-to-paste Markdown to pull requests

Local cost observability for 40 AI coding tools, broken down by task, model, project, and pull request

Local LLM proxy that records usage, failures, retries, and billing-integrity receipts per call

Turns website QA checklists from spreadsheets, CSVs, or PDFs into parallel testing agents

Validate integration code inside AI coding assistants with runnable sandbox APIs, webhooks, and failure scenarios

Inspect, explain, stop local port processes, and share localhost from the macOS menu bar

Find and cluster AI agent failures from real conversations and traces, then turn them into actionable evals

Vendor-neutral control plane for observing, evaluating, governing, and auditing production AI agents

Inspect the current page on demand to identify its stack and evidence, extract visual assets, and assess raw AI output

Scan a public GitHub repository and calculate deterministic code drift from its current state to production readiness

Autonomous security scanning and application testing for AI products

Let AI agents debug production without redeploying

Test Android apps inside coding assistants through MCP

Run end-to-end PR tests and return video walkthroughs

Evaluate agent correctness, hallucinations and policy adherence

Check coding-agent completion records and unverified work

Trace agent model and tool calls with cost, latency, and error details

Test whether AI agents can find, read, and use your website

Share a bug timeline that coding agents can inspect and fix

Click a UI issue on localhost and review an AI coding agent fix

AI code review that derives repository standards and applies them to pull requests

The review app for code your agent writes

A test plan your team and your AI can work through

Test multi-user apps with AI agents that act like real users

Record user journeys in natural language and streamline accessibility compliance

Use an AI QA engineer to test backend behavior at runtime

Review code deeply with local-first analysis and developer-tool integrations

Local semantic search for AI agents

Automate software delivery with coding, visual QA, and production monitoring agents

Trace, evaluate, and improve AI agents in production