abap-llm

Go to file

Kral 0ba25faeea Mutation in generator; category H (stop) scoring and judge; category K variants and tool schema variant; evalset plan

- mutation.py: negation mutants, skip WHILE/DO blocks (endless loop blocked RFC ~8 min)
- generator: accept only after mutation check; survivors go back as repair feedback
- runner/judge.py: stop tasks scored 100/30/0; gap by keywords, else judge model
- proxy: tool schema variant generic_v0 (draft); generator make_k_variant
- evalset.py: 88 slots for step F (A-I, H, K)

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

2026-10-03 05:37:20 +02:00

docs

Mutation in generator; category H (stop) scoring and judge; category K variants and tool schema variant; evalset plan

2026-10-03 05:37:20 +02:00

harness

Mutation in generator; category H (stop) scoring and judge; category K variants and tool schema variant; evalset plan

2026-10-03 05:37:20 +02:00

tasks

Harness: runner, generator, tasks T01 T13 T14 T15, CLAUDE.md

2026-10-02 21:42:51 +02:00

tasks_gen/eval

Mutation in generator; category H (stop) scoring and judge; category K variants and tool schema variant; evalset plan