Repository navigation
evidence(OMN-15241, OMN-13930): independent OCC evidence mint for 5 blocked product PRs - #5171
Conversation
…locked product PRs Append-only mint by an INDEPENDENT verifier (not the implementer of any of the five PRs; no-self-evidence rule). Every probe recorded here was executed by the author verbatim from the committed contract text, and every falsifiability claim was proven by running a RED control -- not asserted. OMN-15241 (5 net-new entries): dod-omnimarket-pr-1919-hook-wired-92986d89 dod-omnibase-compat-pr-183-hook-wired-78ad16e4 dod-omnidash-pr-273-hook-wired-d461b57c dod-occ-pr-5143-hook-wired-9e9ae927 dod-omn-15241-seeded-red-run-30269831694 Each product-PR entry asserts BOTH halves of the wiring at the pinned head -- the .pre-commit-hooks wrapper script AND the .pre-commit-config.yaml entry that invokes it (5143 additionally asserts the ci.yml SKIP-list clause). exit 0 at head, exit 1 at the repo's dev merge base, for all four. The fifth entry re-verifies the seeded-RED proof the ticket rests on: omnimarket run 30269831694 concluded failure on job 'required-check-skip-guard / check-skip-vectors' and the paired clean control run 30264429868 concluded success; throwaway PR #1914 is CLOSED unmerged. OMN-13930 (1 net-new entry): dod-omnibase-infra-pr-2503-deploy-scope-304891d1 Deploy-scope probe over the two runtime paths the deploy-gate flagged. Scope stated honestly: #2503 ships no deployed artifact (no Dockerfile, compose file, k8s manifest, or image build context; `onex delegate` is a local CLI entry point), so the gate's own sanctioned hermetic `gh api .../contents/...?ref=<sha>` form is the probe that ran, not a live-runtime probe that would be vacuous here. Four clauses, the load-bearing one being that cli_delegate.py CALLS check_omnimarket_drift -- an import alone would leave the surface unguarded, which is the exact defect OMN-13930 fixes. exit 0 at 304891d1, exit 1 at dev. Append-only: 6 net-new dod_evidence entries, 6 net-new receipt files, ZERO deletions and zero edits to any existing entry or receipt (git numstat on contracts/: 140 insertions, 0 deletions). Read-only throughout: no deploy, restart, lane mutation, Kafka mutation, database mutation, or production mutation. Evidence-Ticket: OMN-15241
|
Warning Review limit reachedYou’ve reached a temporary PR review limit under our Fair Usage Limits Policy. Next review available in: 25 minutes Enable usage-based reviews in Billing to review now. Otherwise, wait until the next included review is available. How can I continue?After more reviews become available, a review can be triggered using the To avoid repeated limits, reduce automatic review volume by pausing incremental auto-reviews earlier, using label-based review opt-in, excluding WIP or generated PR titles, or requesting reviews manually when the PR is ready. If your team needs uninterrupted high-volume reviews, an organization admin can enable usage-based reviews. How do review limits work?CodeRabbit enforces per-developer PR review limits for each organization. Most developers receive the normal plan review availability. For paid Pro and Pro+ PR reviews, CodeRabbit uses adaptive limits for sustained high-volume activity. When a developer's recent PR review activity reaches the 95th percentile or higher among CodeRabbit users, additional reviews become available more gradually as earlier reviews age out of the rolling window. Please refer docs for additional details. Review details⚙️ Run configurationConfiguration used: defaults Review profile: CHILL Plan: Pro Run ID: 📒 Files selected for processing (10)
✨ Finishing Touches🧪 Generate unit tests (beta)
Comment |
…eipts
Net-new occ-self-bind-pr-5171 entry on both contracts (OMN-14650), plus their
PASS receipts. Probe executed by the author: gh pr view 5171 --repo
OmniNode-ai/onex_change_control --json number,state => {"number":5171,"state":"OPEN"}.
Append-only: additions only, zero deletions.
Evidence-Ticket: OMN-15241
1 participant
What
Independent-verifier OCC evidence mint for five blocked product PRs. Their builders correctly refused to author their own evidence (no-self-evidence rule); this PR is authored by a verifier who built none of them.
Append-only.
contracts/OMN-15241.yamlandcontracts/OMN-13930.yamlalready exist on OCCdev, so this adds net-newdod_evidenceentries and net-new receipt files only — zero deletions, zero edits to any existing entry or receipt.git diff --numstat origin/dev...HEADisN 0on every one of the 8 files.validator_occ_append_only --base-ref origin/devreturnsok: truefor both tickets.OMN-15241 — 5 net-new entries
dod-omnimarket-pr-1919-hook-wired-92986d8992986d89dod-omnibase-compat-pr-183-hook-wired-78ad16e478ad16e4dod-omnidash-pr-273-hook-wired-d461b57cd461b57cdod-occ-pr-5143-hook-wired-9e9ae9279e9ae927dod-omn-15241-seeded-red-run-3026983169430269831694Each product-PR probe asserts both halves of the wiring at the pinned head — the
.pre-commit-hooks/reject-required-check-skip-vector.shwrapper and the.pre-commit-config.yamlentry that invokes it. One clause alone is not evidence: a script with no config entry is dead code, and a config entry with no script is a broken hook — either would satisfy a single-clause probe while leaving the CI-only-enforcement gap OMN-15241 exists to close.#5143carries a third clause for its repo-specificci.ymlSKIPlist change.The fifth entry re-verifies the seeded-RED proof the ticket rests on, rather than taking the ticket's word for it: omnimarket run
30269831694concluded failure with the failing job named exactlyrequired-check-skip-guard / check-skip-vectors, and the paired clean-tree control run30264429868concluded success. The conjunction is what makes it evidence — a RED alone could be ambient breakage. Throwaway PRomnimarket#1914is CLOSED unmerged, so the seed never reacheddev.OMN-13930 — 1 net-new entry
dod-omnibase-infra-pr-2503-deploy-scope-304891d1— the deploy-gate blocker on omnibase_infra#2503.Scope stated honestly, not papered. #2503 ships no deployed artifact: the changed-file set is
docs/runbooks/…md, foursrc/omnibase_infra/cli/*.py, threetests/unit/cli/*.py— no Dockerfile, no compose file, no k8s manifest, no image build context.onex delegateis a locally-installed CLI entry point, so there is no running container,.201lane, or cluster surface whose state adocker/kubectl/rpk/psqlprobe could interrogate for this change; such a probe would be vacuous theatre. The deploy-gate's own failure text names the hermeticgh api …/contents/<changed_path>?ref=<sha>form as the sanctioned probe for exactly this compute-only / no-live-deploy class, and that is what ran.Four clauses over the two runtime paths the gate flagged. The load-bearing one is that
cli_delegate.pycallscheck_omnimarket_drift— an import without a call would leaveonex delegateunguarded, which is precisely the defect OMN-13930 fixes.Falsifiability — proven by execution, not asserted
Every
check_valuewas extracted from the committed contract text (parsed YAML, not retyped) and executed verbatim, then re-executed with the pin swapped to the merge base:For the content probes the RED control is
?ref=dev(the merge base, where the artifacts do not exist). For the seeded-RED entry it is inverting the asserted run conclusion. Every probe reads the product repo over the live GitHub API — none readsdrift/dod_receipts/…orcontracts/OMN-*.yaml, so none can be satisfied by text this PR authored.Local gate readback
yamlfmtreflows these files; parsed semantics were diffed pre/post and are identical on all 8 files plus the 4 pre-existing receipts it touched — so the per-entrycontract_entry_sha256bindings hold.Scope / risk
Evidence-only.
proof_class: receipt-bound. No product source touched, no lane mutation, no runtime touch, no deploy, no restart, no Kafka/DB/prod mutation, no merge. Push originated on.200(stickybeatz-studio) per CLAUDE.md rule 11a, viagit bundle— every file mutation happened in one place (the local worktree) and the pushed commit SHA was verified byte-identical after transfer, so there is no edit-locality mismatch and no vacuous green.Evidence-Ticket: OMN-15241