Seahaven
A framework for building synthetic worlds for AI agents. Useful for RL and evaluation environments.
Kiln-AI/Seahaven
View on GitHub ↗A framework for building synthetic worlds for AI agents. Useful for RL and evaluation environments.
Seahaven grew +0.0% in the last 30 days, ranks 78th of 80 in RL Environments, actively maintained, last commit 3 days ago.
Star trajectory
36 days tracked+0.0% in 30d, below the RL Environments median of +0.7%.
Last commit 3 days ago, fresher than the median of 122d.
How it compares in RL Environments
Ranked 78th of 80 by stars.
Related repos
Ranked by shared tags, then keywords, then starsWhere Seahaven sits
Hover or focus a node for details. Click to open it.
- This repo
- Related repo
- Tag or category
Creating realistic RL environments requires collaboration between researchers, engineers, and domain experts…
+87 (+24%) in 30d · Committed today · Python · Apache-2.0
Shares RL Environments, Evaluation & Guardrails
CommerceAgentBench: Benchmarking Long-Horizon Agents in High-Fidelity, Stateful, and Reproducible Replicas of…
+12 (+0.9%) in 30d · Committed 85d ago · HTML · Apache-2.0
Shares Evaluation & Guardrails, RL Environments
Harness for running and evaluating AI agents against RL environments
+16 (+5.5%) in 30d · Committed today · Python · Apache-2.0
Shares Evaluation & Guardrails, RL Environments
Open source evals for physical AI. Run any LLM/VLA on any arm/humanoid against any real/sim benchmark.
+77 (+11%) in 30d · Committed 10d ago · Python · MIT
Shares Evaluation & Guardrails, RL Environments
ExploitBench measures how far AI agents climb, from reaching vulnerable code, to triggering the bug, to…
+17 (+3.9%) in 30d · Committed 141d ago · Python · MIT
Shares RL Environments, Evaluation & Guardrails
An agentic robotics benchmark for coding agents: 1,156 robot tasks in 35 suites on 47 robots, each answered…
Committed today · Python · MIT
Shares RL Environments, Evaluation & Guardrails
[Public preview] Externally scored agentic ML research benchmark: 60 tasks, real competition ground truth…
+0 (+0.0%) in 30d · Committed 58d ago · Python · Apache-2.0
Shares RL Environments, Evaluation & Guardrails
Trace-native CI/CD for AI agents — production failures become regression tests that block the PR. Auto-detect…
-413 (-27%) in 30d · Committed yesterday · Python · MIT
Shares Evaluation & Guardrails
OrcaReplay — Time travel for AI agents. Record, replay, fork, and debug any agent run with any model. Built by…
+833 (+296%) in 30d · Committed 51d ago · TypeScript · Apache-2.0
Shares Evaluation & Guardrails
Trace-native CI/CD for AI agents: grade production traces, cluster failures, freeze them into replayable…
-305 (-26%) in 30d · Committed 5d ago · Python · MIT
Shares Evaluation & Guardrails
- Language
- Python
- License
- MIT License
- Contributors
- 3
- Forks
- 4
- Watchers
- 0
- Open issues
- 0
- Releases
- 0
- Created
- 2026-09-11
- Stars this week
- +0
You've seen where Seahaven stands today.
Get the moves before everyone else: the 5 RL Environments repos climbing fastest, every Tuesday.
Free weekly digest. Double opt-in, unsubscribe anytime.