You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
SDK provides **code execution proxies** - tools for AI agents. You add it to your toolbox in Vercel AI SDK, Langchain or OpenAI Agents, making LLM write Python or Bash code to talk with Slack or Linear API. Requests will automatically be intercepted and routed to isolated test environments. This enables agents to interact with service replicas without any code changes. See more in: **[Python SDK](sdk/agent-diff-python/README.md)**
150
150
151
151
152
+
## Benchmark & Training
153
+
154
+
-**HuggingFace Dataset**: [hubertmarek/agent-diff-bench](https://huggingface.co/datasets/hubertmarek/agent-diff-bench) — 224 tasks across all 4 services (80/20 train/test split, stratified by service)
155
+
-**Prime Intellect Environment**: [agent-diff-bench on Prime Lab](https://app.primeintellect.ai/dashboard/environments/hubert-marek/agent-diff-bench) — run evaluations or RL training via Hosted Training
156
+
-**Paper**: [AgentDiff: Agentic API Evaluation via State Differencing (KDD 2026 pre-print)](https://drive.google.com/file/d/1BlmJTSMX7ohwvD1aYBByg7_Y815fgsxp/view?usp=sharing)
157
+
152
158
## Evaluations & Test Suites
153
159
154
160
Collections of test cases with assertions that you can run against agent runs using evaluations.
0 commit comments