git-pull-rebase-trap
Verdict: CANT_TELL_YET
- Cut sub-reason
- absent from this receipt
- Unmeasured sub-reason
- absent from this receipt
- Value class
- trap-discipline
- Wrong instrument
- no
- Declared synthetic control
- no
git-pull-rebase-trap sized paired run, n = 32, run 0700d089a0275ed27d3e219a680e1959bd69ec11e7841a4280d95a4e17243907, decided 2026-09-03: both arms passed every epoch (lattice both_pass 32, full_only 0, null_only 0, both_fail 0; signed delta 0.000, 95% CI [-0.107, 0.107]), and the trap-discipline verdict is CANT_TELL_YET. Zero of 32 Null epochs and zero of 32 Full epochs ran `git pull`; all 64 fetched and merged instead, so the armed `pull.rebase=true` configuration was never exercised. The ancestry oracle checks whether history was preserved, and it cannot distinguish an avoided hazard from a hazard the run never entered. Under the skill-harness#403 ruling (2026-09-03), this run carries no information about the trap-discipline estimand: it measured whether the subject rebases by habit, not whether the card's guard changes an outcome. skill-harness#421 makes `evaluate-paired` refuse a run in this shape by name, as HAZARD_NOT_MET, so this receipt records the last read taken under the pre-#421 instrument.
Cost beside evidence
| Leg | Figure | Detail |
|---|---|---|
standing_tokens | REFUSED (not_instrumented) | absent from this receipt |
fired_tokens | REFUSED (not_instrumented) | The body loaded in 24 of 32 Full epochs; token attribution per skill was not measured for this run. |
aux_tokens | REFUSED (not_instrumented) | The body loaded in 24 of 32 Full epochs; token attribution per skill was not measured for this run. |
Clause-level evidence grade
Clause evidence: UNMEASURED (no_extraction: clause evidence requires an extraction output; produce one with skill init --out <file>)
Measurements
Keys the standard defines but this receipt does not carry are listed as absent rather than filled in.
| Key | Figure | Detail |
|---|---|---|
Null-arm (no-skill) pass rate from a Stage-0 screen.p0 | absent from this receipt | absent from this receipt |
Full-arm pass rate from a paired run.full_pass_rate | 1.0 (32/32 epochs) | Every Full epoch also fetched and merged instead of pulling, and every one passed the ancestry oracle. |
Null-arm pass rate from a paired run (may equal p0 when only a screen ran).null_pass_rate | 1.0 (32/32 epochs) | Every Null epoch fetched and merged instead of pulling, and every one passed the ancestry oracle. |
Posterior P(rate above threshold) from the paired fit.p_win | absent from this receipt | absent from this receipt |
Paired per-epoch discordance rate d-hat (and optional CI in detail).discordance_rate | 0.0 (0/32 epochs) | Lattice: both_pass 32, full_only 0, null_only 0, both_fail 0. Signed delta 0.000, 95% CI [-0.107, 0.107]. |
Pre-stated GO/NO-GO apparatus gate when one was registered.go_nogo | NOT_APPLICABLE | absent from this receipt |
Null-arm hazard-entry rate: the fraction of Null epochs where the hazard pattern was entered.hazard_entry_null | absent from this receipt | absent from this receipt |
Full-arm hazard-entry rate: the fraction of Full epochs where the hazard pattern was entered.hazard_entry_full | absent from this receipt | absent from this receipt |
Null-arm completion rate: the fraction of Null epochs where the completion oracle scored pass.null_completion_rate | absent from this receipt | absent from this receipt |
Full-arm completion rate: the fraction of Full epochs where the completion oracle scored pass.full_completion_rate | absent from this receipt | absent from this receipt |
Silent violation rate: the fraction of epochs where completion held but invariant failed (C=1, I=0).silent_violation_rate | absent from this receipt | absent from this receipt |
Value delivery
The skill body was read in addition to the standing description.
pi_c: 0.75 (24/32 trials)
exposure: 1.0 (32/32 epochs)
Evidence admissibility
Status: admissible
Ingested 2026-09-03 under the oracle metric identity 0.4.1 (#416), declared after #411 and #413 edited subject/ingest.py. The re-ingest wrote one runs row from the eval logs on disk, with the runner block (rat_id, skill_id, task_family, estimand, route) recorded verbatim under config_json.
Instrument identity
Figures produced under different instrument identities are not comparable, so the generation stamp travels with the figures.
- extractor_model
anthropic/claude-sonnet-5- prompt_fingerprint
sha256:2289ee87959b603b5a249430efbe80f998e16831b75772ebc99c5d725c80338c- schema_fingerprint
command_succeeds+oracle_template:sha256:9d4101935cccf2d67620e7d580ede092fbaf2bc71a6ecde02cb066b67526139b
Subject identity
- skill_id
387989fed2eda3b083ea6ce8094ff0c3685e08e738b6ed58df97da8242dbcb96- harness_version
0.2.3- metric_version
0.4.1- implementation_hash
ae28a10512c62a5b16a2ca272d07a81510afba9085197e926c50e39c879d09a4- subject_model
absent from this receipt- arms
null, full
Currentness
- state
STALE- basis
NONE- last_checked_at
- absent from this receipt
- next_check_due
- absent from this receipt
- max_age_days
- absent from this receipt
Source of record
- Repo-relative path to the prose source of record.
prose_path - docs/ratifications/RAT-0001-git-pull-rebase-trap.md
- ISO date of the underlying result when known.
date notes- Run 0700d089a0275ed27d3e219a680e1959bd69ec11e7841a4280d95a4e17243907, sized paired run n=32 under RAT-0001, launched 2026-09-03. docs/sers/receipts/superseded/gitpull-paired-k8-2026-09-01-detector-v2.json is the GO datum this run was sized on. Ingest recorded the oracle metric identity as 0.4.1 (#416) after #411 and #413 edited subject/ingest.py; evaluate-paired was fixed to read the runner-declared identity rather than the card content digest (#417). Eval logs at logs-stage2-sized/ under the batch-1 gitpull directory.