Skip to content

Improve README and add opt-in Jev workflows - #252

Open
Coding-Dev-Tools wants to merge 68 commits into
mainfrom
codex/jev-pro-team-pr-20261003-final
Open

Coding-Dev-Tools wants to merge 68 commits into
mainfrom
codex/jev-pro-team-pr-20261003-final

Conversation

@Coding-Dev-Tools

@Coding-Dev-Tools Coding-Dev-Tools commented Oct 3, 2026 •

Copy link
Copy Markdown
Owner

Summary

Adds opt-in dashboard evidence review and experimental BYOK route selection, and moves detailed configuration out of the README. A remote advisory timeout, busy worker, or malformed response preserves the configured local routes within the original planner deadline. Local calls remain available while a remote worker is occupied.

Ask submissions snapshot the question, scope, classification, and consent together. Changing Jev settings invalidates pending responses and retries, so revoked consent cannot be reused. Dashboard evidence review reports whether selected evidence can answer a question; it does not certify a declarative claim's truth.

Managed and auto configurations stay local for route selection and never silently switch to BYOK. Fixed dashboard review workflows resolve their lazy managed client independently after input and consent checks. Classic, Smart, direct MCP, and raw managed custom requests still fail locally before credential refresh.

The changes also retain opt-in compact previews and reject oversized or invalid Unicode payloads before credential lookup. Benchmark labels use the selected retrieval depth and withhold packed scores when no samples were measured. Integration docs match the current local-agent trust policy and Pi installation command.

Current main is integrated, including explicit cross-agent session resume and the MCP SDK pin. Shipped schemas and skill checksums are regenerated. New immutable offline evidence binds all 585 source files; README benchmark links point to the commit containing its guide and matching SVG/PNG visuals. Historical artifacts remain preserved.

Verification

  • Full local offline suite: 7,892 passed, 132 platform/optional-dependency skips.
  • All eight documented retrieval, retention, security, grounded, code-arm, and evidence evaluation gates passed. The three public fixture measurements are unchanged.
  • Ruff, Pyright 1.1.414, commercial manifest, strict-CSP drift, MCP contract, source bindings, and diff checks passed.
  • Browser verification: 20 Ask/Home scenarios, including revoked consent and late-response regressions, and 17 memory-workflow scenarios passed.
  • Prime-agent integration: 149 passed, 1 optional skip.
  • Pi typecheck, 23 package tests, pack inspection, and one real MCP integration test passed. Lock and installed dependency audits report zero vulnerabilities after the documented test-host leaf repair.
  • All 29 GitHub checks, including CI and CodeQL, passed on head de04e64552c549e2464ce1119a12c1c57931b044. The coverage job passed 8,040 tests with 20 skips and 86.00% coverage; the complete browser job passed 194 scenarios.

Managed service boundary

Managed availability remains not_yet_available pending private service acceptance and capacity qualification. The documented limits are per eligible individual: 100 evaluated questions per rolling hour, 1,000 per rolling five hours, and 2,000 per rolling 24 hours. Questions count individually; command review uses two. Direct BYOK can incur provider charges. The documented fleet-cap conflict must be resolved before managed activation. No held-out quality, cost-saving, activation, or deployment claim is made.

@chatgpt-codex-connector

chatgpt-codex-connector Bot commented Oct 3, 2026 •

Copy link
Copy Markdown

Codex Review Summary

This comment shows the latest Codex review activity on this pull request.

Review Status Commit Review trigger
📝 Code Review ✅ Completed 2026-10-07T09:57:52.949629Z de04e64 New commits
ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

  • Open a pull request for review
  • Mark a draft as ready
  • Comment "@codex review" or "@codex security review".

Codex reacts with 👀 while any review is running, comments if it has suggestions, and reacts with 👍 once all reviews finish with no findings.

@chatgpt-codex-connector chatgpt-codex-connector Bot left a comment

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: 71c099dfd8

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

  • Open a pull request for review
  • Mark a draft as ready
  • Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

Comment thread engraphis/routes/v2_api.py Outdated
Comment thread engraphis/mcp_server.py

@chatgpt-codex-connector chatgpt-codex-connector Bot left a comment

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: 9165f7847f

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

  • Open a pull request for review
  • Mark a draft as ready
  • Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

Comment thread engraphis/backends/jev_decision.py

@chatgpt-codex-connector chatgpt-codex-connector Bot left a comment

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: 3bec5bc6ed

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

  • Open a pull request for review
  • Mark a draft as ready
  • Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

Comment thread docs/benchmark-evidence/offline-fixtures-v131.json

@chatgpt-codex-connector chatgpt-codex-connector Bot left a comment

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: 24f0fd4973

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

  • Open a pull request for review
  • Mark a draft as ready
  • Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

Comment thread tests/test_documentation_contracts.py Outdated

@chatgpt-codex-connector chatgpt-codex-connector Bot left a comment

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: cf6b4268ae

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

  • Open a pull request for review
  • Mark a draft as ready
  • Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

Comment thread engraphis/core/recall.py Outdated
Comment thread engraphis/factory.py

@chatgpt-codex-connector chatgpt-codex-connector Bot left a comment

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: d299855513

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

  • Open a pull request for review
  • Mark a draft as ready
  • Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

Comment thread engraphis/dashboard_assets/ledger.js Outdated

@chatgpt-codex-connector chatgpt-codex-connector Bot left a comment

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: 6ab0475f13

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

  • Open a pull request for review
  • Mark a draft as ready
  • Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

Comment thread engraphis/backends/jev_decision.py Outdated
Comment thread eval/jev_recall_quality.py Outdated

@chatgpt-codex-connector chatgpt-codex-connector Bot left a comment

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: 0f6d292beb

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

  • Open a pull request for review
  • Mark a draft as ready
  • Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

Comment thread engraphis/backends/jev_decision.py

@chatgpt-codex-connector chatgpt-codex-connector Bot left a comment

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: 94b8d244bf

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

  • Open a pull request for review
  • Mark a draft as ready
  • Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

Comment thread engraphis/core/recall.py
Comment thread engraphis/service.py Outdated

@chatgpt-codex-connector chatgpt-codex-connector Bot left a comment

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: fd5b060adb

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

  • Open a pull request for review
  • Mark a draft as ready
  • Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

Comment thread docs/benchmark-evidence/offline-fixtures-v148.json

@chatgpt-codex-connector chatgpt-codex-connector Bot left a comment

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: 062bff700d

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

  • Open a pull request for review
  • Mark a draft as ready
  • Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

Comment thread engraphis/factory.py Outdated

@chatgpt-codex-connector chatgpt-codex-connector Bot left a comment

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: c76873dc06

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

  • Open a pull request for review
  • Mark a draft as ready
  • Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

Comment thread tests/test_benchmark_evidence.py Outdated
PUBLIC_OFFLINE_ARTIFACT = "offline-fixtures-v128.json"
PUBLIC_OFFLINE_SHA = "73f2d1a8cd6e2db070577582a2266f6605efc052800da17db9938f7a274bf755"
PUBLIC_OFFLINE_ARTIFACT = "offline-fixtures-v149.json"
PUBLIC_OFFLINE_SHA = "d5d36c55c4303d77b161137521dd31f77f39b7a0c9e2fed3ddb63e303b12cc6d"

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

P1 Badge Regenerate the v149 evidence checksum

Fresh evidence beyond the earlier v148 finding shows that the newly pinned v149 artifact itself hashes to 23b5d2bf01f2…, while this constant, its .sha256 sidecar, BENCHMARKS.md, and the generated visuals all claim d5d36c55c430…. Consequently, python -m pytest tests/ -q -x fails in test_public_numeric_evidence_registry_is_complete_and_live, blocking the required offline gate; regenerate or repin the v149 artifact and update every checksum consumer.

AGENTS.md reference: AGENTS.md:L50-L54

Useful? React with 👍 / 👎.

@chatgpt-codex-connector chatgpt-codex-connector Bot left a comment

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: d1bd37d783

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

  • Open a pull request for review
  • Mark a draft as ready
  • Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

Comment thread docs/MCP_TOOLS.md

Copy link
Copy Markdown
Owner Author

One historical integrity item remains on e21e321.

The retained docs/benchmark-evidence/offline-fixtures-v149.json is missing its final LF. Its stored bytes hash to 23b5d2bf01f2bd6ef82f3991c908a7cc9987f13779671c948244117c71ff7caf. Adding exactly one final LF yields d5d36c55c4303d77b161137521dd31f77f39b7a0c9e2fed3ddb63e303b12cc6d, matching the existing sidecar and advertised historical identity. Please restore only that byte, preserve all JSON data, and confirm the current editing branch/HEAD checkpoint. Keep the v149 thread open until the stored checksum pair is verified.

The other three findings can be closed with evidence:

  • Managed route planning: factory.py:215–228 selects no managed client; the managed/auto regression forbids credential lookup.
  • Direct managed custom calls: mcp_server.py:2683–2689 rejects them before client selection; tests cover direct, Classic and Smart dispatch.
  • v148 live-registry complaint: the retained artifact checksum matches its sidecar, and its recorded recall/service hashes match introduction commit 0bd6a8e. It is a valid historical snapshot; the active registry now uses offline-fixtures-managed-preflight-review-20261006.json. No v148 rewrite is needed.

All 26 CI jobs and CodeQL passed on this head. No repeated benchmark run is needed to restore the missing v149 byte.

Integrate current main, keep managed fixed reviews independent from BYOK-only planning, validate payload bytes before credential lookup, retain opt-in compact previews, and correct selected-report labels and integration docs. Regenerate immutable source-bound evidence and shipped contracts.

Copy link
Copy Markdown
Owner Author

Read-only handoff for de04e64552c549e2464ce1119a12c1c57931b044. All 26 CI jobs and CodeQL pass on this head. The new Ask/UI, payload-validation and rendering fixes should be preserved. Four bounded items remain:

  • Historical v149 still lacks its final LF: stored SHA256 23b5d2bf01f2bd6ef82f3991c908a7cc9987f13779671c948244117c71ff7caf; adding exactly one LF produces advertised d5d36c55c4303d77b161137521dd31f77f39b7a0c9e2fed3ddb63e303b12cc6d. Preserve all JSON data and keep the checksum thread open until the stored pair is verified. A retained-hash regression should prevent recurrence.
  • engraphis/backends/jev_decision.py:261 accepts a successful response returned after its deadline. An exact-head synthetic probe returns decision at 101 ms under a 100 ms budget. Recheck the deadline after evaluate; timely results must still work and late results must fall back without reordering local routes.
  • engraphis/routes/v2_api.py:1530–1537 replaces an explicitly injected offline_mode=True, client=None backend with a managed client. Preserve explicit offline backends and injected clients without checking credentials before consent; replace only the ordinary managed/auto planning stub. Test both modes and denied consent.
  • .github/workflows/release.yml:1292–1296 uploads assets, then promotes using only the earlier Latest result. Re-read and validate Latest immediately before promotion. Regression: candidate v1.7.9, initial Latest v1.7.8, external v1.8.0 during upload must not promote; malformed/failed lookups must fail closed. Keep queue:max and qualification/waiver gates. This narrows the race; it is not an atomic lease.

Please have the current editor incorporate these small fixes, run focused regressions, and refresh source-bound evidence/pins with a new immutable artifact when Python source changes. Please confirm the editing branch and HEAD checkpoint before another writer publishes.

This branch has not been deployed

No deployments
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant