EveryDev.ai
Subscribe
Home
Tools

5,089+ AI tools

  • New
  • Trending
  • Featured
  • Rate tools
  • Compare
  • Arena
Categories
  • Agents2594
  • Coding1910
  • Infrastructure898
  • Projects718
  • Research640
  • MCP611
  • Marketing497
  • Analytics495
  • Design422
  • Testing406
  • Data378
  • Security328
  • Integration259
  • Communication250
  • Prompts246
  • Extensions219
  • Commerce171
  • Voice170
  • DevOps168
  • Web123
  • Learning114
  • Finance35
AI Tools by Topic
  • AI Coding Assistants
  • Agent Frameworks
  • MCP Servers
  • AI Prompt Tools
  • Vibe Coding Tools
  • AI Design Tools
  • AI Database Tools
  • AI Website Builders
  • AI Testing Tools
  • LLM Evaluations
Follow Us
  • X / Twitter
  • LinkedIn
  • Reddit
  • Discord
  • Threads
  • Bluesky
  • Mastodon
  • YouTube
  • GitHub
  • Instagram
Get Started
  • Users
  • Rate Tools
  • About
  • Editorial Standards
  • Corrections & Disclosures
  • Community Guidelines
  • Advertise
  • Contact Us
  • Newsletter
  • Submit a Tool
  • Start a Discussion
  • Write A Blog
  • Share A Build
  • Terms of Service
  • Privacy Policy
Explore with AI
  • ChatGPT
  • Gemini
  • Claude
  • Grok
  • Perplexity
Agent Experience
  • llms.txt
Theme
With AI, Everyone is a Dev. EveryDev.ai ©
    1. Home
    2. Tools
    3. AetherTrace
    AetherTrace icon

    AetherTrace

    Observability Platforms
    Featured

    A local-first, open-source visual debugger, reverse proxy and replay engine for tracing and debugging multi-agent LLM workflows.

    Visit Website

    At a Glance

    Pricing
    Open Source

    Free, MIT-licensed, local-first agent execution workstation with no subscription or hosted backend required.

    Engagement

    Available On

    CLI
    Web
    SDK
    API

    Resources

    WebsiteDocsGitHubllms.txt

    Topics

    Observability PlatformsMulti-agent SystemsLLM Evaluations

    Alternatives

    LaminarArize AICoze Loop
    Developer
    ZLouadaThe limits of the dead

    Listed Oct 2026

    About AetherTrace

    AetherTrace is an open-source, local-first workstation for debugging AI agent execution, published by ZLouada under the MIT license. It runs a reverse proxy that records LLM calls into a local SQLite database and shows them in a browser-based Flight Deck. The project is currently at v0.3.

    What It Is

    AetherTrace sits between an agent application and its model provider. You point your SDK's baseURL at the local proxy (http://localhost:4000/v1), and it captures agents, models, tools, retries, HTTP requests and errors as a live execution DAG with a timeline and span inspector. It works with OpenAI, Anthropic, Ollama or local models, and is started with npx aethertrace start, a Docker image, or from source.

    Debugging, Breakpoints and Replay

    Live breakpoints suspend matching requests so you can inspect state, mutate prompts or tool arguments, return a mock response, or abort. Runs can be forked from any span, reusing cached prefixes so unmutated ancestors cost nothing, with cache-only, safe or live replay modes (live replay requires explicit confirmation). A chaos engine can inject latency, synthetic 429 and 502 errors and malformed JSON. Circuit breakers enforce budget caps and stop repeated-request loops.

    Experiments, Evaluation and Diffing

    Experiment Studio creates variants from existing traces by changing prompts, models, tool arguments or mocks, then batch-replays them. Evaluation Studio checks cost, latency, tokens, reliability and topology with structured assertions, clusters errors and exports redacted JSON or Markdown reports. Diff Arena aligns two runs semantically to show cost, latency, token and lineage differences. The CLI assert command can act as a CI regression gate.

    Sharing and Integrations

    Runs export as portable .trace.json bundles that teammates can import into their own instance. Provenance capture records Git, dependency and environment context. OpenTelemetry OTLP/HTTP traces can be imported, and LangChain and LangGraph callback handlers plus TypeScript and Python SDKs preserve parent-child relationships. Data stays local, and secret redaction and SSRF protection are included.

    AetherTrace - 1

    Community Discussions

    Be the first to start a conversation about AetherTrace

    Share your experience with AetherTrace, ask questions, or help others learn from your insights.

    Pricing

    OPEN SOURCE

    Open Source

    Free, MIT-licensed, local-first agent execution workstation with no subscription or hosted backend required.

    • MIT license
    • Local-first SQLite storage
    • Zero-code-change proxy
    • Execution DAG and span inspector
    • Live interception, fork and replay

    Capabilities

    Key Features

    • Live execution DAG and timeline with span inspector
    • Drop-in reverse proxy for OpenAI-compatible calls
    • Live breakpoints with prompt mutation and mock responses
    • Fork and replay with cached prefixes
    • Experiment Studio for variant replays
    • Evaluation Studio with structured assertions
    • Semantic Diff Arena
    • Chaos fault injection
    • Budget and loop circuit breakers
    • Portable .trace.json export and import
    • Secret redaction and SSRF protection
    • Git and environment provenance
    • CI regression gate via CLI assert
    • Local SQLite (WAL) storage

    Integrations

    OpenTelemetry
    LangChain
    LangGraph
    OpenAI
    Anthropic
    Ollama
    TypeScript SDK
    Python SDK
    Docker
    GitHub Actions
    API Available
    View Docs

    Ratings & Reviews

    No ratings yet

    Be the first to rate AetherTrace and help others make informed decisions.

    Rate other tools you’ve used

    Developer

    ZLouada

    The limits of the dead

    Read more about ZLouada
    WebsiteGitHubLinkedInX / Twitter
    1 tool in directory

    Similar Tools

    Laminar icon

    Laminar

    Open-source platform to trace, evaluate, and analyze AI agents with real-time observability and powerful evaluation tools.

    Arize AI icon

    Arize AI

    Arize AI is an enterprise AI and agent engineering platform for development, observability, and evaluation of LLM applications, AI agents, and ML models in production.

    Coze Loop icon

    Coze Loop

    Open-source platform for developing, debugging, evaluating, and monitoring AI agents across their full lifecycle.

    Browse all tools

    Related Topics

    Observability Platforms

    Comprehensive platforms that combine metrics, logs, and traces with AI-powered analytics to provide deep insights into complex distributed systems and application behavior.

    160 tools

    Multi-agent Systems

    Platforms for creating and managing teams of AI agents that can collaborate.

    511 tools

    LLM Evaluations

    Platforms and frameworks for evaluating, testing, and benchmarking LLM systems and AI applications. These tools provide evaluators and evaluation models to score AI outputs, measure hallucinations, assess RAG quality, detect failures, and optimize model performance. Features include automated testing with LLM-as-a-judge metrics, component-level evaluation with tracing, regression testing in CI/CD pipelines, custom evaluator creation, dataset curation, and real-time monitoring of production systems. Teams use these solutions to validate prompt effectiveness, compare models side-by-side, ensure answer correctness and relevance, identify bias and toxicity, prevent PII leakage, and continuously improve AI product quality through experiments, benchmarks, and performance analytics.

    201 tools
    Browse all topics
    Back to all toolsSuggest an edit
    ratings
    discussions