Skip to content

Spec: CodexHub 0.1.8 staged compatibility and release roadmap #248

Open
5 of 9 issues completed
Open
Spec: CodexHub 0.1.8 staged compatibility and release roadmap#248
5 of 9 issues completed
Labels
P1Priority 1: current iteration goalcampaign:0.1.80.1.8 release train campaign mapenhancementNew feature or requestready-for-agentFully specified, ready for an AFK agentwayfinder:mapWayfinder campaign map
Milestone
0.1.8 — Responses, Chat & CLI stability

Description

@NOirBRight
NOirBRight
opened on Jul 29, 2026

Current verification policy (2026-08-04)

GitHub Actions CI is disabled for this project. Beta/RC/stable work is verified locally on the exact candidate: focused tests, affected full suites, CLI/manual checks, and the documented local release checklist. Historical Hosted runs remain evidence of past candidates only and are not a current merge or release gate.

Status

0.1.8-beta.3.3 is the current published prerelease.

Beta3.1–3.3 retained the Beta3 advanced-tool release while adding the verified reasoning/history, Kimi Chat translation, message-phase, and context-derived output-limit corrections. Existing task files are not rewritten and no Provider/model fallback was added.

The active frontier is now Beta4. The remaining train separates:

  1. exact runtime-derived generic Collaboration V2 over Responses;
  2. three-route maintained Ollama Responses regression;
  3. usable Ollama Chat plus complete Collaboration V2 on those same three routes;
  4. stability freeze, fixed RC batches, and human stable promotion.

Product contract

Model selection and route identity

  • The user selects the Provider and model.
  • CodexHub resolves that choice to one immutable endpoint/auth/protocol/model binding.
  • Tool compatibility never selects a model.
  • Alias, catalog mismatch, missing capability, adaptation, retry, or error handling must never fallback to Terra, Luna, Official, or another Provider/model.

Tool compatibility

Each request derives compatibility from the installed Codex runtime declaration type and the already-selected upstream protocol:

  • native — preserve the compatible declaration and wire lifecycle unchanged.
  • adapt — use an Adapter only for an incompatible but reversible structural shape.
  • omit — do not expose an optional unavailable/non-representable tool.
  • required-but-unavailable — fail visibly before sampling when the request requires that tool.

Already compatible tools are not routed through an Adapter. Namespace/custom aliases are request-scoped and reversible. The Gateway adapts protocol shapes but does not execute tools, create agents, schedule Collaboration work, or fabricate results.

Hosted tools are exposed only when the selected Provider natively supports them. Explicit and implicit cross-Provider tool proxying are prohibited.

Unknown custom endpoints use the same conservative type/protocol rules. GLM-5.2, K2.7 Code, and DeepSeek V4 Flash 0731 are maintained-Provider regression samples, not an exhaustive allowlist or runtime eligibility gate.

Release train

0.1.8-beta.1 — usability hotfix (released)

Gate #258.

Shipped urgent Official-continuation and provider-scoped context-window fixes, exact RoutePlan/K2.7 foundations, local verification migration, release artifacts, and regressions required to make the application usable again.

0.1.8-beta.2 — generic tool compatibility core

Gate #318.

Outcome: custom Responses endpoints receive one generic native/adapt/omit/required-but-unavailable path with exact model identity and no cross-Provider proxy. Beta2 adds no alternate-codec retry or model-output format repair.

0.1.8-beta.3 — generic advanced tools, official Collaboration controls, and Ollama stream diagnosis

Gate #281.

Outcome: add no global V1/V2 switch and no model fallback. Official catalog values remain the baseline; only an accepted, exact-model user override changes the effective value and survives restart/sync/catalog rebuild. V2 selectors are shown only for list-visible Official models with a complete accepted matrix verdict. Generic client tool discovery and Code Mode remain model-agnostic. Ollama transport failures are classified separately from semantic/tool incompatibility; no blanket timeout increase or blind replay is accepted.

0.1.8-beta.4 — runtime-derived generic Collaboration V2

Gate #284.

Outcome: complete the generic spawn/message/follow-up/wait/interrupt/list lifecycle over custom Responses routes using the actual runtime V2 direct-tool surface. Native-compatible shapes remain native; only complete reversible lifecycles adapt. Essential non-representable semantics fail before sampling. Codex Client remains the agent owner/executor; Gateway performs no scheduling and does not rewrite task history.

0.1.8-beta.5 — three-route maintained Ollama Responses regression

Gate #351.

Outcome: independently regression-test every generic Responses capability shipped in Beta2 through Beta4 against all three Provider/model configurations maintained by CodexHub. Beta5 adds no planned generic capability, model-specific codec, runtime allowlist, or fallback.

0.1.8-beta.6 — usable Ollama Chat plus complete Collaboration V2

Gate #259.

Outcome: all three maintained Ollama Chat routes complete progressive SSE, multi-turn text, standard function tools, default reasoning, classified errors, a real read/apply_patch/verify loop, and the complete client-owned Collaboration V2 lifecycle without hidden Responses/Official fallback or V2-to-V1 downgrade. Chat Code Mode/custom/freeform, client tool_search, arbitrary namespace, search, and hosted parity remain separately dispositioned and are not implied by V2 support.

0.1.8-beta.7 — stability and release-evidence freeze

Gate #319.

This is the final planned beta and adds no capability.

0.1.8-rc.1 — six fixed exact-candidate regressions

Gate #257.

These are release regressions for maintained Provider configurations, not runtime restrictions on user-configured endpoints. Any product-code change after the fixed batches invalidates the RC candidate.

Stable 0.1.8 — promotion only

Gate #320.

Stable promotion requires the accepted RC lineage, bounded maintainer smoke, no open 0.1.8 P0/P1 regression, exact asset/updater readback, and a human go/no-go. Product-code changes after RC invalidate the candidate.

Current frontier

The active sequence is:

Beta3 issue-state reconciliation
        ↓
#392 exact runtime/wire contract
        ↓
#199 configuration safety + #282 implementation
        ↓
#283 real CLI/App/fixture verification
        ↓
#252 → #284 (Beta4)
        ↓
#275 + #276 + #393 → #351 (Beta5 Responses)
        ↓
#66 → #394 → #395
        ↓
#285 + #286 + #396 → #253/#254 → #259 (Beta6 Chat + V2)
        ↓
#319 → six RC batches → #257 → #320

#324 remains research input rather than a Beta4 runtime allowlist. Broad third-party route matrices, catalog-backed subagent rosters, cross-Home topology/history migration, and Chat Code Mode parity remain post-0.1.8 unless separately promoted.

Mandatory release order

Every beta and RC follows:

  1. implementation complete;
  2. exact candidate SHA review accepted;
  3. CLI/manual verification accepted;
  4. local verification gate succeeds on that lineage;
  5. tag and prerelease publication.

A failed or incomplete gate stops publication. CI success alone never replaces review or CLI/manual acceptance.

Explicitly out of scope

  • Any explicit or implicit cross-Provider tool proxy.
  • Runtime model qualification records or named-model allowlists.
  • Model selection by CodexHub.
  • Provider-hosted capability emulation.
  • Gateway-owned tool execution or a second agent scheduler.
  • Beta2 alternate-codec retries or model-output format repair.
  • Chat Code Mode/custom/freeform, client tool_search, arbitrary namespace, and hosted parity beyond the accepted Beta6 contracts.
  • Anthropic Messages/Claude Code integration, provider-onboarding redesign, complete cross-Home history migration, and non-Windows packaging.

Done when

  • Beta1 usability prerelease is published and read back.
  • Beta2 publishes the generic tool compatibility core.
  • Beta3 publishes generic client tool discovery and reversible Code Mode, official model-level V1/V2 controls, and Ollama stream/error classification; Beta3.1–3.3 hotfix regressions remain cumulative.
  • Beta4 validates the exact runtime-derived generic Responses Collaboration V2 lifecycle without V1 corruption.
  • Beta5 validates the complete Responses surface against GLM-5.2, K2.7 Code, and DeepSeek V4 Flash 0731.
  • Beta6 validates usable Ollama Chat and the complete Collaboration V2 lifecycle against the same three routes.
  • Beta7 closes stability/evidence work and freezes the RC input.
  • RC1 passes all six fixed maintained-Provider batches on one exact candidate.
  • Stable 0.1.8 is promoted only through 0.1.8 stable gate: Promote the accepted RC lineage #320.

Activity

  1. NOirBRight
    added
    ready-for-agentFully specified, ready for an AFK agent
    on Jul 29, 2026

  2. NOirBRight commented on Jul 29, 2026

    @NOirBRight
    NOirBRight
    on Jul 29, 2026
    OwnerAuthor

    Decision update: retain only necessary CLI E2E evidence

    The 0.1.8 qualification runner must not retain full real-client work roots after a completed run.

    • The durable release evidence is limited to the CLI qualification summary, exact candidate/model/route manifest, bounded per-case outcome records, hashes, and sanitized diagnostics required to explain a failure.
    • Successful-case isolated Codex homes, client state, copied runtimes, repeated candidate builds, VM images, GUI seeds, and complete work directories are transient and are deleted during runner finalization.
    • A failed case may retain only a bounded diagnostic bundle selected by the runner; retaining the complete isolated tree is not an accepted fallback.
    • 0.1.8 keeps no Desktop/ZCode/manual qualification state. The single durable product seam remains the real Codex CLI profile agreed in this specification.
    • Cleanup must run on success, failure, timeout, and cancellation, and the summary must report whether transient-root cleanup completed.
  3. NOirBRight
    added this to the 0.1.8 — Responses, Chat & CLI stability milestone on Jul 29, 2026
  4. NOirBRight
    added
    enhancementNew feature or request
    wayfinder:mapWayfinder campaign map
    P1Priority 1: current iteration goal
    on Jul 29, 2026
  5. NOirBRight
    added a parent issue on Jul 29, 2026
  6. NOirBRight
    added sub-issues on Jul 29, 2026

  7. 43 remaining items

  8. NOirBRight
    added a sub-issue on Jul 31, 2026

  9. NOirBRight commented on Aug 1, 2026

    @NOirBRight
    NOirBRight
    on Aug 1, 2026
    OwnerAuthor

    Beta1 completion update (2026-08-02)

    The active stop-the-bleeding Beta1 gate #258 is complete and closed.

    The historical core-Responses Beta1 section below is retained as design context; its qualification graph is now Beta2 as stated at the top of this issue. Next frontier: review and plan the deferred Beta2 tickets (#62, #65, #58, #249, #273, #274, #275, #276) without claiming their capabilities in Beta1.

  10. NOirBRight
    added sub-issues on Aug 2, 2026

  11. NOirBRight commented on Aug 2, 2026

    @NOirBRight
    NOirBRight
    on Aug 2, 2026
    OwnerAuthor

    This was generated by AI during triage.

    2026-08-02 full roadmap and issue audit

    The post-Beta1 roadmap has been reconciled without changing the approved 0.1.8 capability direction.

    Release-train changes

    Important inserted fixes

    Triage corrections

    Current open allocation: 46 in 0.1.8, 13 in 0.1.9, 8 in 0.1.10, and 19 unmilestoned backlog/operational items.

    Immediate executable frontier

  12. NOirBRight commented on Aug 2, 2026

    @NOirBRight
    NOirBRight
    on Aug 2, 2026
    OwnerAuthor

    Beta2 insertion: catalog override ownership (#327). New P1 issue #327 preserves explicit same-model catalog edits such as Luna multi_agent_version=v2 across restart/sync without changing the pinned Official default (Luna remains v1). It also preserves user-owned custom model_catalog_json paths and prevents Official planner metadata from leaking into third-party rows. #327 may be developed while #62 tooling CI runs, but must merge before authorized #62 live evidence is bound to the release catalog and before #318; the existing #62 -> #65 -> #249 -> #275/#276 -> #58 path remains unchanged.

  13. NOirBRight
    added a sub-issue on Aug 4, 2026
  14. NOirBRight
    changed the title [-]Spec: CodexHub 0.1.8 staged prerelease and qualification roadmap[/-] [+]Spec: CodexHub 0.1.8 staged compatibility and release roadmap[/+] on Aug 4, 2026
  15. NOirBRight
    added
    campaign:0.1.80.1.8 release train campaign map
    on Aug 4, 2026

  16. NOirBRight commented on Aug 4, 2026

    @NOirBRight
    NOirBRight
    on Aug 4, 2026
    OwnerAuthor

    Campaign grouping and Beta2 orchestration update (2026-08-04)

    The 0.1.8 roadmap now uses one campaign label per release stage while retaining the aggregate 0.1.8 milestone:

    Beta2 dispatch order

    #62 + #350
   ↓
#64
   ↓
#65
   ↓
#198
   ↓
#249 (maintainer decision)
   ↓
#58
   ↓
#318 (release gate)

    #327 is already merged and must remain in the Beta2 candidate lineage before the authorized #62 live evidence is bound. #350 is already merged. PR #355 is the current #62 implementation candidate at exact head 61949b36c44174a8fa15b4665d7645ec70c68a7a; its independent Spec and Quality reviews are in progress. It remains unqualified because the retained real wire capture is 0.144 reference data and the 0.146 source contract is explicitly not_observed/unqualified.

    Each executable ticket will be handled by one isolated Luna Max worker, then exact-SHA review, CLI/manual validation, and Hosted CI / gate before serial integration. #249 and publication gates remain maintainer/coordinator decisions; a worker must not make those product or release decisions.

    No #65 dispatch, Beta2 candidate tag, or Beta2 release is allowed until #62's current acceptance gates are actually observed and reviewed. Every prerelease follows: implementation → exact-SHA review → CLI/manual validation → Hosted CI / gate → tag/release.

  17. NOirBRight commented on Aug 4, 2026

    @NOirBRight
    NOirBRight
    on Aug 4, 2026
    OwnerAuthor

    CI blocker update (2026-08-04)

    Hosted run 30904259353 on the #62 candidate exposed two unrelated Rust debug lifecycle fixture failures. They are tracked in new P1 #356 and are being handled by a separate Luna Max worker with test-only scope. #356 must produce a new exact candidate with focused 10x normal/debug repetitions and a full Hosted matrix; no retry of the failed job is accepted. This does not change #62's evidence contract or unlock #65.

  18. NOirBRight commented on Aug 4, 2026

    @NOirBRight
    NOirBRight
    on Aug 4, 2026
    OwnerAuthor

    Decision: local verification only (2026-08-04)

    Per maintainer direction, GitHub Actions CI is disabled for CodexHub (CI workflow state: disabled_manually). The CI / gate required status check was removed from both dev and main; branch protection itself remains in place.

    From this point through Beta2, Beta3, Beta4, Beta5, Beta6, Beta7, RC1, and stable:

    1. implement on an isolated worker branch;
    2. perform exact-SHA Spec/Quality review;
    3. run local focused and affected full suites, CLI/manual checks, and the documented local release checklist;
    4. record the local verification result;
    5. only then merge/tag/publish according to the campaign gate.

    Existing Hosted run links are historical evidence only. No ticket may wait for or claim a Hosted CI / gate result.

  19. NOirBRight commented on Aug 7, 2026

    @NOirBRight
    NOirBRight
    on Aug 7, 2026
    OwnerAuthor

    Beta3 plan sync: reasoning-stream outcome

    Beta3 fixes the Gateway's third-party reasoning-progress regression without exposing raw chain-of-thought. Standard third-party Responses reasoning-summary events are downstream progress, while raw chain-of-thought and provider-local encrypted state remain hidden and non-forwardable. A valid slow stream must be delivered as completion; slow-valid streams, upstream failures, incomplete SSE, client cancellation, and Gateway deadline failures remain distinct classifications. No blanket timeout increase, blind replay, cross-Provider proxy, or model fallback is permitted.

    Plan: docs/superpowers/plans/2026-08-07-beta3-advanced-responses-and-reasoning.md

  20. NOirBRight commented on Aug 10, 2026

    @NOirBRight
    NOirBRight
    on Aug 10, 2026
    OwnerAuthor

    Approved roadmap update: three Ollama routes and Chat Collaboration V2 in Beta6

    The maintainer approved two scope changes:

    1. Beta5 maintained Responses regression now covers three Ollama Cloud routes: GLM-5.2 (0.1.8 Beta5: Run the full maintained GLM-5.2 Provider regression #275), K2.7 Code (0.1.8 Beta5: Run the full maintained K2.7 Code Provider regression #276), and DeepSeek V4 Flash 0731 (0.1.8 Beta5: Run the full maintained DeepSeek V4 Flash 0731 Responses regression #393).
    2. Complete third-party Chat Collaboration V2 moves from 0.1.9 into 0.1.8 Beta6. Beta6 now uses one generic implementation/verification pair (0.1.8 Beta6: Implement generic Collaboration V2 over Chat Completions #394/0.1.8 Beta6: Verify the complete generic Chat Collaboration V2 lifecycle #395) and qualifies the Ollama Chat endpoints for GLM-5.2 (0.1.8 Beta6: Complete the GLM-5.2 Ollama Chat and V2 CLI path #285), K2.7 Code (0.1.8 Beta6: Complete the K2.7 Code Ollama Chat and V2 CLI path #286), and DeepSeek V4 Flash 0731 (0.1.8 Beta6: Complete the DeepSeek V4 Flash 0731 Chat CLI path #396).

    The cumulative RC therefore adds fixed DeepSeek Responses and Chat batches (#397/#398), producing six fixed batches under #257. The 0.1.9 milestone retains Chat Code Mode/custom/freeform/tool_search/arbitrary-namespace/hosted parity, broad matrices, cross-Home V2 migration, and other post-0.1.8 expansion; it no longer owns the three-route Chat V2 promise.

    This does not authorize model-name production branches, fallback, cross-Provider execution, Gateway agent scheduling, or disk-history rewriting.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

No one assigned

    Labels

    P1Priority 1: current iteration goalcampaign:0.1.80.1.8 release train campaign mapenhancementNew feature or requestready-for-agentFully specified, ready for an AFK agentwayfinder:mapWayfinder campaign map

    Projects

    No projects

      Milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions