AgentDiff
Trajectory regression testing for AI agents
What it does
Traditional evals test outputs, missing when agents loop on tools or burn 50% more tokens. AgentDiff fixes this by recording agent runs as DAGs and diffing execution trajectories against golden baselines directly in CI. It catches tool loops, cost spikes, and latency regressions before merge evaluating how the model arrived at the answer, not just what it returned. Framework-agnostic with adapters for OpenAI Agents, Langfuse, LangSmith, and OpenInference.
AgentDiff compares agent execution traces as DAGs in CI/CD - automatically detecting trajectory drift, redundant tool loops, and cost regressions.
Your agent returned the right answer, but took 5 extra steps and burned 3× the tokens. AgentDiff catches silent tool loops, cost surges, and execution drift in CI before your PR lands in production. Traditional assertions only check if the final output string matched. They are blind to execution drift — missing recursive tool loops, unverified prompt detours, and 3× cost surges before code merges. The agent still delivered the expected summary, so standard assertion tests pass easily in CI. Behind the scenes, a prompt tweak caused the agent to loop 3 times over the customer database. AgentDiff detects the step surge in milliseconds and fails the PR before it ever costs you real money. No…from agentdiff.lostmartian.in
Does the same job
all alternatives →
Progress AI Observability30d ago · telerik.com · ▲168Trace, evaluate, and improve AI agents in production
More ai this month
the category →
I trained a 125M-parameter transformer to autocomplete piano performances in real time (~108 notes/sec on an iPhone 15). The idea is basically GitHub Copilot or Tabnine, except instead of prompting it with code, you prompt it by playing a few notes on a MIDI piano. The model then continues what you played, entirely on-device. The app is free if anyone wants to try it. Happy to answer questions about the model, training, Core ML, or the many things that didn't work.
AI · 16d ago · simedw.com
Astute▲585Automate your B2B brand going viral, with new media creators
AI · 18d ago · company-app.joinastute.com
Hey HN, Henry from Cactus here! We previously released Cactus Needle, a 14MB agentic LLM for tool call, device use, and structured extraction for phones, wearables, smart homes, small robots and microcontrollers. We got really great feedback here, and have now incorporated the suggestions to release Needle 2. The whole model is a single 14MB binary that runs a full session in 28MB of RAM; 45m parameters at 2bit compression. Needle hits 500 tokens/sec decode speed on a Raspberry Pi 5, sits between 400-1,500 tokens/sec on VR devices like Meta Quest 3S and Apple Vision Pro, and ranges…
AI · 26d ago · cactuscompute.com
Launched alongside, August 2026
the whole month →- TL
Life & fun · 10d ago · louisabraham.github.io
- SA
Hello HN! I found that picking out plausible but diverse skin tones for my digital art and game development projects was kind of difficult, and I got curious about if there was a way to define a color space that made it easy. I've built a color picker and procedural generation algorithm based on the space as well as a bunch of other fun js features and demos throughout the page that use the equations. If you find it interesting, I have lots of explanations of how I built it and what properties the space has. The methodology might be a bit shaky, but hopefully the result is as helpful for…
Life & fun · Aug 2026 · toneyalexander.github.io
I trained a 125M-parameter transformer to autocomplete piano performances in real time (~108 notes/sec on an iPhone 15). The idea is basically GitHub Copilot or Tabnine, except instead of prompting it with code, you prompt it by playing a few notes on a MIDI piano. The model then continues what you played, entirely on-device. The app is free if anyone wants to try it. Happy to answer questions about the model, training, Core ML, or the many things that didn't work.
AI · 16d ago · simedw.com