Decision Workspace
nornir-codegen-judge vs onejudge vs solve-score
Side-by-side comparison of Rust crates
The compile-and-test codegen JUDGE for the nordisk constellation — a pure-Rust (no-LLM) answer evaluator. Drops an already-produced code answer (a full file body or a unified diff) into a scratch copy of a target crate, runs `cargo build` + an acceptance command, and turns the outcome into a 0.0..=1.0 quality `Verdict`. The `AnswerJudge` trait is the shared accept/reject oracle reused by the bake-off scorer AND the funnel task-decomposition loop; `CodegenJudge` is the first impl. Warehouse-free, self-contained (extracted from nornir per the shrink plan).
A simulated interaction and evaluation loop over oneharness: drive a harness through a multi-turn conversation and score the transcript.
Independent implementation of SOLVE v0.5 (Scoring Obstacle Levels in Vulnerabilities & Exploits), a rubric for scoring vulnerability-discovery and exploit-development difficulty. Reproduces the 16 published calibration assessments. Not affiliated with or endorsed by Irregular.
Core Metrics
| nornir-codegen-judge | onejudge | solve-score | |
|---|---|---|---|
| Health Score | 43 | 60 | 49 |
| Total Downloads | 71 | 13.6K | 18 |
| 30d Downloads | 0 | 0 | 0 |
| Dependents | 3 | 220 | 0 |
| Releases | 1 | 43 | 1 |
| Last Updated | 66d ago | 2d ago | 58d ago |
| Age | 2m | 3m | 1m |
Health Breakdown
Technical Details
| nornir-codegen-judge | onejudge | solve-score | |
|---|---|---|---|
| Version | 0.1.0 | 0.18.1 | 0.1.0 |
| Stable (≥1.0) | ✗ No | ✗ No | ✗ No |
| License | MIT OR Apache-2.0 | MIT | MIT OR Apache-2.0 |
| Dependencies | 3 | 10 | 2 |
| Crate Size | 10KB | 290KB | 26KB |
| Features | 0 | 3 | 2 |
| Yanked % | 0.0% | 0.0% | 0.0% |
| Edition | 2024 | 2021 | 2024 |
| MSRV | — | 1.89 | 1.85 |
| Owners | 1 | 1 | 1 |
Links
Quick Verdict
- •onejudge leads with a health score of 60/100, but none of the options score above 80.
- •onejudge is depended on by 220 crates — strongest ecosystem trust.