Skip to content

Find code by asking what it does. A CLI for coding agents that uses Jev to discover relevant files and source context.

docs: record Django confirmation and current cohort handoff

David Zhang8cb0688ccf1d
diff --git a/specs/jevgrep/README.md b/specs/jevgrep/README.md
index be8299b..2c2773d 100644
--- a/specs/jevgrep/README.md
+++ b/specs/jevgrep/README.md
@@ -7,72 +7,44 @@ Jev classification through Vercel AI Gateway. This is the build plan; production
## Next Agent Prompt
You are implementing this spec. Status: **implementing**, updated **2026-09-26**.
-Next: collect the running Django attempt in `installed-jg-cpython-work-clock-v1`,
-then grade/account and apply the registered Django-first continuation rule.
-The [agent-work timing protocol](assets/timing-protocol.md) passed review and
-merged Docker checks; all ten cells are frozen and no-call validated.
-The [source placement experiment](assets/presentation-study.md) is terminal and
-failed; its remaining nine cells stay unrun. Keep production renderer and skill unchanged.
-The no-model counterfactual preserves all evidence while moving first-ranked
-source inside the observed tool prefix. Real SQLite replays match the frozen oracle, and an 81-request
-three-file replay proves both pipelines exceed the follow-up threshold and skip
-that pass. Exact coverage and whole-pipeline limits are in the
-[mechanism-first skill experiment](assets/query-framing-study.md). That candidate
-failed the same Django regression at Sol $1.2353512; its remaining nine tasks stay
-unrun and the production skill is unchanged. Packet-visibility comparison is
-complete: initial truncation hid selected source, but a later read and explicit
-recognition preceded replacing the failing regression with a type-changing test.
-Do not infer another query-breadth fix. Django under the work-time protocol is
-the only active paid task; do not launch a replacement attempt.
-The frozen `installed-jg-cpython-parity-v1` cohort is complete and not accepted:
-7/10 solves versus 8/10 baseline, six cost wins, and 35.27% lower full Sol cost.
-[Confirmation evidence](assets/cpython-confirmation.md) owns all outcomes, trace
-findings and quality caveats. Never rerun a baseline, replace an attempt, or treat
-the lower total as compensation for a lost solve. Keep the measured CLI unchanged
-while testing any query/skill hypothesis as a separately identified candidate.
-
-The merged `bun run verify` gate passed and whole-product review's sole finding
-is fixed with a clean follow-up review. Exact archive checks pass on macOS arm64
-and Linux arm64/amd64. [Runtime evidence](assets/python-runtime.md) records the
-artifact and interpreter limits. The architecture port is verified for the
-exercised corpus; task-quality acceptance failed and remains required.
-
-`installed-jg-reference-parity-v1` is stopped and superseded. Requests and
-scikit-learn officially solved but exceeded baseline cost; Django was interrupted
-with its patch and traces retained; seven cells never started. These superseded
-studies remain stopped. The bundled-CPython cohort is terminal. The completed
-query diagnostics made only bounded Jev calls; the separate mechanism-first plan
-owns the new skill candidate and its Django-first evaluation.
-
-The changed-policy cohort `installed-jg-final-cohort-v1` was stopped and marked
-superseded, retaining completed outcomes and interrupted traces. It is not a
-release acceptance cohort. Never rerun saved baselines or pool candidates.
-Report Jev observed API costs separately from scored full Sol cost.
-
-Keep the production incumbent's restored expansion, discovery inputs/order,
-excerpt presentation, scoped guidance and accepted skill (only rename executable
-to `jg`) unchanged. The isolated experiment changes only the recorded skill
-instruction and does not promote that change into production. Required
-Node-only runtime, auth, cache, filesystem protections and truthful failure
-handling remain, but must not silently change healthy retrieval decisions.
-The earlier deterministic and platform gates passed for the deviating port;
-they do not prove parity. See [parity restoration](assets/parity-restoration.md) and the
-[source-budget investigation](assets/source-budget-decision.md).
-Remaining: a candidate that passes frozen quality confirmation, final choices
-consolidation, whole-spec closeout review and close-spec.
-See [contracts](contracts.md) for the corrected contract.
-
-No user decision blocks starting. External verification dependencies are a working
-Docker daemon, explicit Gateway credentials for live tests, retained official task
-snapshots/baselines, and macOS access for native install verification. Missing live
-inputs do not block deterministic development, but do block the corresponding
-acceptance claim. Do not repair credentials or regenerate an existing baseline.
-
-Implement one focused slice at a time, running its narrow tests and recording its
-artifact. Use write-tests for behavioral tests and review before completing an
-implementation pass. Update this section, slice status, decisions and unresolved
-verification before ending each pass. The named commands are implemented; their
-recorded evidence does not substitute for the open quality gate.
+The only active confirmation is `installed-jg-cpython-work-clock-v1`.
+Django officially solved below its immutable baseline; the registered rule now
+permits the other nine tasks in that same frozen plan. The sequential runner has
+started scikit-learn. Collect every terminal receipt, official grade and complete
+Sol bill, then aggregate the full cohort. Never replace an attempt or rerun a
+baseline. [Timing protocol and current evidence](assets/timing-protocol.md) own
+this study's procedure and results.
+
+Keep production retrieval and the canonical skill unchanged. The installed
+package matches the restored bundled-CPython candidate; only the prospective
+benchmark timing protocol differs. The failed query and presentation experiments
+have not earned promotion. Their remaining cells stay unrun. Prior outcomes
+cannot be pooled into the active cohort.
+
+Evidence and boundaries:
+
+- [Runtime verification](assets/python-runtime.md): merged default tests,
+ whole-product review, and exact archive checks on macOS arm64 and Linux
+ arm64/amd64 passed. Interpreter parity is established for the exercised corpus,
+ not all possible inputs.
+- [Restoration record](assets/parity-restoration.md): frozen reference, known
+ deviations and controlled request/output/recovery comparisons.
+- [Completed first confirmation](assets/cpython-confirmation.md): rejected for
+ losing a baseline solve despite lower total cost. Its traces remain evidence.
+- [Query trial](assets/query-framing-study.md) and
+ [presentation trial](assets/presentation-study.md): rejected experiments,
+ source-parity probes and limits on causal explanations.
+
+Priority: finish this frozen quality confirmation; address any failing contract
+without weakening acceptance; run whole-spec closeout review; consolidate choices
+against the final code; then close-spec. Slice 08 remains open. No publication or
+release tag is authorized for the intentional `0.0.0` development checkpoint.
+Report observed Jev costs separately; they never enter scored Sol task cost.
+
+No user decision blocks continuing. Use the authorized credentials and retained
+Docker snapshots; do not repair credentials or regenerate baselines. Preserve
+user-edited skill files and local historical evidence. Stage only explicit files,
+keeping deprecated personal-repository evals out of Git.
- [x] [01 — Reference and real HTTP fixture](slices/01-reference.md)
- [x] [02 — Bundled parser parity](slices/02-parser.md)
diff --git a/specs/jevgrep/assets/parity-restoration.md b/specs/jevgrep/assets/parity-restoration.md
index 703ebea..8d42db8 100644
--- a/specs/jevgrep/assets/parity-restoration.md
+++ b/specs/jevgrep/assets/parity-restoration.md
@@ -4,9 +4,9 @@ This record preserves the pre-CPython restoration work and its superseded
cohorts. The bundled-CPython port subsequently passed integration, platform and
whole-product checks; [runtime evidence](python-runtime.md) owns that coverage.
Its completed quality cohort failed acceptance, as recorded in
-[confirmation evidence](cpython-confirmation.md). The separate
-[mechanism-first skill experiment](query-framing-study.md) owns the active trial;
-none of the stopped studies below is reactivated.
+[confirmation evidence](cpython-confirmation.md). The separate query and presentation experiments also failed.
+[The prospective timing protocol](timing-protocol.md) owns the active faithful
+confirmation; none of the stopped studies below is reactivated.
The user rejected unvalidated architecture changes during production porting.
The accepted reference is the source closure and skill bound by
diff --git a/specs/jevgrep/assets/timing-protocol.md b/specs/jevgrep/assets/timing-protocol.md
index 7bc4ae2..f519694 100644
--- a/specs/jevgrep/assets/timing-protocol.md
+++ b/specs/jevgrep/assets/timing-protocol.md
@@ -71,7 +71,18 @@ Study `installed-jg-cpython-work-clock-v1` is frozen and all ten no-call checks
pass. Its 179 installed files match the original bundled-CPython cohort exactly;
the skill hash is
`1a6b29ca8855d3538883c6dada289db68dd2dfca90f466e39407902dc000dd50`.
-Proof: `/tmp/jg-work-clock-freeze-proof.json`. Django is running first under the
-registered rule. Collect its terminal receipt, official grade and complete bill
-before any continuation. No other cell has started. Production CLI and canonical
-skill remain unchanged.
+Proof: `/tmp/jg-work-clock-freeze-proof.json`.
+
+Django completed with valid timing evidence and officially resolved (one solve,
+zero infrastructure/errors). All 37 Sol requests are accounted for: **$1.1409928**
+versus the saved **$1.5055472** baseline. Jev's known observed subtotal is
+**$0.148011822**; its complete total remains unknown (776 cost responses for 777
+client calls) and is excluded from scoring.
+
+The receipt records 359.286 seconds wall time, 222.354 charged work and 136.933
+credited retrieval. This run fits the old 900-second wall limit too: its success
+must not be attributed to gaining more time. Query and agent trajectory vary.
+The official pass and lower complete Sol bill satisfy the registered continuation
+rule, so the sequential runner has started the remaining nine tasks in this same
+plan. No Django replacement or baseline execution occurs. Production CLI and
+canonical skill remain unchanged; full-cohort acceptance is still unproven.
diff --git a/specs/jevgrep/slices/08-release.md b/specs/jevgrep/slices/08-release.md
index 77d1cc0..8d835ec 100644
--- a/specs/jevgrep/slices/08-release.md
+++ b/specs/jevgrep/slices/08-release.md
@@ -1,6 +1,6 @@
# Confirm quality and the release artifact
-Status: frozen confirmation complete and rejected; solve preservation fails on Django and only six cost wins are verified. Depends on: 07. Read [contracts](../contracts.md) first.
+Status: open. The first faithful confirmation was rejected; the new frozen work-clock cohort is running after its Django-first continuation gate passed. Depends on: 07. Read [contracts](../contracts.md) first.
## Contract and owner
@@ -28,8 +28,9 @@ One frozen release-candidate dossier.
[the runner workflow](../../../evals/implementation/swebench/installed.md); native
macOS installed smoke command documented by the harness.
-Commands are implementation targets. Add them in this slice; do not imply they
-already exist. CLI transcripts replace visual/screenshot gates for this product.
+These commands are implemented. Use the frozen installed runner to finish the
+open quality gate and retain its evidence. CLI transcripts replace visual/screenshot
+gates for this product.
## Verification and verdict