Null resultAI usage (tokens)
Rule checks from conditional prohibitions, the rest of the band
Does the one paying rule-check shape repeat on the candidates not yet measured?
measured-local2026-09-1980e4a13
- docs/experiments/DETECTION-2026-09-19.md · 80e4a13 · 2026-09-20
Headline figures
5
Skills attempted
measured-local · 2026-09-19
2of 5
Reached a three-arm comparison
measured-local · n=5 · 2026-09-19
0of 2
Accepted
95% interval 0.0% to 65.8%
measured-local · n=2 · 2026-09-19
-29.0%
Mean token reduction over compared skills; negative is dearer
measured-local · 2026-09-19
2of 2
Compared skills whose repetitions disagree in sign
measured-local · n=2 · 2026-09-19
2
Stopped as artifact_refused
measured-local · 2026-09-19
1
Stopped as tasks_not_discriminating
measured-local · 2026-09-19
Notes
- The conditional band has 20 candidates in the measurable corpus; lane K measured 15 and these are the other 5.
- Two were refused by the mission rule before measurement: a commit skill whose rule check forbids committing, and a publish skill whose rule check forbids pushing.
Tables
n: Runs per arm behind the row: repetitions times tasks compared, or the graded runs of the eligibility check when the skill was not compared. 5 rows.
| skill | name | shape | installs | status | status_superseded | decision | accepted | artifacts | skill_md_unchanged | token_reduction | spread_min | spread_max | straddles_zero | net_call_change | tokens_none | tokens_original | tokens_optimized | score_none | score_original | score_optimized | mission_preserved | retention_band | R | tasks | n | evidence_class | date | commit |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| review-pr | Review PR | workflow | — | compared | — | inconclusive_repetitions_straddle_zero | No | hook | Yes | 0.0768 | -0.1468 | 0.0758 | Yes | 0 | 91,707 | 105,046 | 96,983 | 0.5 | 0.75 | 1 | Yes | full | 3 | 2 | 6 | measured-local | 2026-09-19 | 80e4a13fb96ebb96d8efff8690a8c77215657590 |
| dmr-from-django-ninja | DMR from django-ninja | workflow | 45 | compared | — | inconclusive_repetitions_straddle_zero | No | hook | Yes | -0.6561 | -4.0859 | 0.5735 | Yes | 5 | 53,954 | 396,121 | 656,023 | 0 | 0.8 | 0 | No | lost | 3 | 1 | 3 | measured-local | 2026-09-19 | 80e4a13fb96ebb96d8efff8690a8c77215657590 |
| atmos-secrets | Atmos Secrets | reference-lookup | 5 | artifact_refused | — | — | No | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 3 | 1 | 3 | measured-local | 2026-09-19 | 80e4a13fb96ebb96d8efff8690a8c77215657590 |
| publish-project-to-github | Publish Project to GitHub | workflow | — | artifact_refused | — | — | No | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 3 | 2 | 6 | measured-local | 2026-09-19 | 80e4a13fb96ebb96d8efff8690a8c77215657590 |
| terraform | Terraform IaC Expert | workflow | — | tasks_not_discriminating | — | — | No | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 3 | 2 | 6 | measured-local | 2026-09-19 | 80e4a13fb96ebb96d8efff8690a8c77215657590 |