Overview & Result
Decouples tool decision-making from code generation. Jev decides whether to run tests, lint, git diff, or build commands, leaving Codex unburdened to focus on code synthesis.
How Jev fits in the loop
- Analyze current repo state and terminal errors.
- Jev selects appropriate developer command.
- System executes tool and pipes output into Codex.
How to reproduce
- Inspect the source implementation at https://github.com/micic-mihajlo/jev-tool-runner.
- Configure local Jev API credentials and register required tool/model bindings.
- Execute test prompts and observe decision latency and routing outputs.
Why this build matters
Prevents LLMs from getting lost in circular tool selection arguments.
Limitations
- Metrics are author-reported from the initial release unless independently verified.
- Requires access to the respective agent framework or runtime environment.
- Generative model execution remains external to the Jev decision step.