harnessx
HarnessX’s historical-best acceptance gate, as a HarnessGrad method.
The method returns one edited harness per round. It never sees a score for the harness it is producing — the platform measures every candidate on the same study set and records the result in curve.jsonl.
Measured results
Section titled “Measured results”Scores, rounds and per-split comparisons are on the ALFWorld benchmark page — one table for every method, so the same number cannot say two different things in two places.
Record
Section titled “Record”This method’s rounds on the Hub — each round’s harness, the edit that produced it, and its per-task trajectories: open the directory.
Deviations from the method it ports
Section titled “Deviations from the method it ports”5 deliberate difference(s), recorded in methods/harnessx/method.json:
- Rollback restores the staged historical-best tree and declares incumbent_round, but grading is delayed until the next method call.
- The upstream shipping validator is not executed; version selection alone does not implement its preflight checks.
- The optional cost penalty uses platform tokens rather than upstream USD.
- The passed-count guard has a legacy aggregate fallback and accepts breaches with unavailable counts.
- Edit generation uses the shared editor, with JSON pre-registration rather than the resident HarnessX meta-agent and journal workflow.
methods/harnessx/run.py and its rule test in tests/quality/. python3 tools/method_lint.py prints one line per method.