Repository Issue Activity (beta)

#1881 feat(sandbox): enroot sandbox provider

#987 Can you recompute reward using existing rollout and a new verifier?

#1256 feat: agent skill evaluation infrastructure

#1640 Feat: Add Orchestration Layer for Evaluation

#1338 epic: dataset preparation

#1542 [epic] Expand benchmark test coverage as part of benchmark migration in Nemo Gym

#1451 Epic: Validated Benchmarks to NeMo Gym for Nemotron

#1606 Uniform judge-failure reporting in the metrics file

#1866 Epic: Gym architecture discussion regarding components, modularity and environments

#1867 Feature: standardized trajectories format

#1863 Epic: Standardized result files and output artifacts

#1582 [Epic] Search & Discoverability for Environments and Components from the CLI

#1938 test item

#1912 nemo gym scoring parser bug fix

#1935 Can you recompute reward using existing rollout and a new verifier?: Add implementation

#1934 Can you recompute reward using existing rollout and a new verifier?: Prepare RFC

#1909 openai_model 500s on gpt-oss hosted-MCP (mcp_call) output items, aborting rollout collection

#1593 Inspect-AI Based Gyms

#340 docs: how to use vllm extra body

#625 feat: running `ng_run` outside of the gym repository

#936 bug: ng_init_resources_server crashes when using relative paths or used outside Gym root

#973 [BUG?] Function calling environment

#1919 Shared internal OpenSandbox endpoint and local Enroot provider

#1494 epic: agent skill improvement

#1685 epic: built-in sandbox providers