Skip to content

docs(opencode-provider): soft-deprecate in favor of @omniroute/opencode-plugin (#3419) - #3613

Merged
diegosouzapw merged 1 commit into
release/v3.8.21from
fix/3419-opencode-deprecate
Jun 11, 2026
Merged

docs(opencode-provider): soft-deprecate in favor of @omniroute/opencode-plugin (#3419)#3613
diegosouzapw merged 1 commit into
release/v3.8.21from
fix/3419-opencode-deprecate

Conversation

@diegosouzapw

@diegosouzapw diegosouzapw commented Jun 11, 2026

Copy link
Copy Markdown
Collaborator

Closes #3419 (the deprecation-decision open question; the RFC's two code PRs #3417 + #3418 already merged)

Decision

Option A — soft-deprecate. @omniroute/opencode-provider writes a static model list to opencode.json that drifts behind the live OmniRoute catalog; @omniroute/opencode-plugin fetches /v1/models at OpenCode startup and stays live. The provider package is kept working (no code/behavior change) but is no longer the recommended path.

Changes

  • package.json — npm description now leads with DEPRECATED — use @omniroute/opencode-plugin instead.
  • README.md — prominent deprecation banner at the top with the one-line opencode.json migration ("plugin": ["@omniroute/opencode-plugin"]).
  • Guard test pins the deprecation notice (description + README) so it can't be silently dropped.

No production runtime code touched — the package still builds and emits valid config.

Validation

@omniroute/opencode-provider test suite: 53/53 green (incl. the new #3419 guard); package.json re-parses clean.

Note: the registry-side npm deprecate @omniroute/opencode-provider "..." flag is applied at publish time (not in-repo) — to be run with the next release of the package.

Co-authored-by: herjarsa herjarsa@users.noreply.github.com

@diegosouzapw @herjarsa
docs(opencode-provider): soft-deprecate in favor of @omniroute/openco…
a8958fe 
…de-plugin (#3419)

Co-authored-by: herjarsa <herjarsa@users.noreply.github.com>
@gemini-code-assist

gemini-code-assist Bot commented Jun 11, 2026

Copy link
Copy Markdown
Contributor

Warning

You have reached your daily quota limit. Please wait up to 24 hours and I will start processing your requests again!

@diegosouzapw
diegosouzapw merged commit 5c11d1c into release/v3.8.21 Jun 11, 2026
2 checks passed
@diegosouzapw diegosouzapw mentioned this pull request Jun 11, 2026
Closed
diegosouzapw added a commit that referenced this pull request Jun 11, 2026
@diegosouzapw @herjarsa
@oyi77 @claude @dhaern
Release v3.8.21 (#3593)
c315a23 
* chore(release): open v3.8.21 development cycle

* fix: pass through valid max_tokens-truncated responses instead of fake 502 (#3572) (#3595)

* fix: /v1/completions returns legacy text-completion format, not chat (#3571) (#3596)

* fix: z.ai/GLM coding plan no longer shows Monthly 0% when no monthly cap (#3580) (#3597)

* docs: mark DISCOVERY_TOOL_DESIGN endpoints as Phase-2 not-yet-implemented (#3498) (#3599)

* fix(agent-bridge): add validate-only upstream-ca/test route (#3488) (#3600)

* fix(gamification): add level/badges/badges-earned profile routes (#3484)

* security(oauth): migrate 5 public client_ids to resolvePublicCred (#3493)

* fix(mcp): ship MCP server source closure in npm files + coverage gate (#3578)

* fix: add reasoning token buffer for combo routing (fixes #3587) (#3588)

Integrated into release/v3.8.21

* Refactor: Extract chatCore phases into modular files (#3598)

Integrated into release/v3.8.21 — chatCore phase modularization. Adjusted: re-derive idempotencyKey for the save path after the check moved into the module (co-authored). Thanks @oyi77!

* docs(changelog): credit #3598 (chatCore modularization) + #3588 (combo reasoning buffer)

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>

* fix(api): implement GET /api/guardrails + POST /api/guardrails/test, drop shadow/guardrails doc-fiction (#3496) (#3602)

Integrated into release/v3.8.21 — implements GET /api/guardrails + POST /api/guardrails/test, removes shadow/guardrails doc-fiction. TDD-validated (5/5) + check-docs-symbols/typecheck/eslint green.

* fix(gemini): isolate textual reasoning wrappers (#3605)

Split-out PR C from #3584. Isolates textual reasoning wrappers (<think>/<thinking>/<thought>/<internal_thought>, including malformed/open tags) into reasoning_content across both the non-streaming sanitizer and the Gemini streaming translator, with split-chunk buffering. Additive to the existing textual tool-call pipeline; does not touch the #3569 native functionResponse path. Integrated into release/v3.8.21. Thanks @dhaern!

* fix(antigravity): normalize Gemini 3.5 Flash tier IDs (#3603)

Split-out PR A from #3584. Normalizes the Antigravity/agy Gemini 3.5 Flash tier IDs to clean public names (gemini-3.5-flash-low/medium/high), maps them to the live upstream IDs at the executor boundary, and removes Antigravity from the global model resolver so the executor owns wire normalization. Maintainer follow-up: kept gemini-3.5-flash-preview as a hidden backward-compat alias routing to the High tier (so saved combos/configs keep working). Live-validated the tier set via the agy CLI catalog. Integrated into release/v3.8.21. Thanks @dhaern!

* fix(agent-bridge): surface real MITM startup-failure cause, not always port 443 (#3606) (#3608)

Integrated into release/v3.8.21 (#3606)

* fix(oauth): surface real Kiro import-token failure cause, not a bare 500 (#3589) (#3609)

Integrated into release/v3.8.21 (#3589)

* docs(opencode-provider): soft-deprecate in favor of @omniroute/opencode-plugin (#3419) (#3613)

Integrated into release/v3.8.21 (#3419)

* fix(usage): normalize Antigravity and agy provider quotas (#3604)

Split-out PR B from #3584. Normalizes Antigravity/agy provider quotas: prefers retrieveUserQuota for live consumption, falls back to fetchAvailableModels and local usage_history, sanitizes cached Provider Limits so retired upstream IDs are not re-exposed, and schedules a deduplicated post-usage refresh. Maintainer follow-up: decoupled the post-usage refresh via a lightweight usageEvents bus (usageHistory no longer dynamic-imports providerLimits) so it does not pull the executors/translator graph into the typecheck-core surface — typecheck:core stays at 0. Integrated into release/v3.8.21. Thanks @dhaern!

* feat(cli): add autostart on/off/toggle shorthand for headless serve mode (#3331) (#3614)

Integrated into release/v3.8.21 (#3331)

* docs(changelog): credit #3603 (Flash tier IDs) + #3604 (provider quotas) + #3605 (reasoning wrappers)

Co-authored-by: diegosouzapw <diegosouza.pw@gmail.com>

* fix(review): resolve findings from /review-reviews battery (v3.8.21 hardening) (#3618)

Pre-release hardening from the /review-reviews battery — 15 findings resolved (L1-L13,L15) + L14 live-verified WONTFIX, convergence re-review clean. lint/typecheck:core/test:vitest(146)/build green; zero new test:unit failures vs baseline 797de43.

* chore(release): v3.8.21 CHANGELOG + i18n + env-doc sync

---------

Co-authored-by: Hernan Javier Ardila Sanchez <hjasgr@gmail.com>
Co-authored-by: Paijo <14921983+oyi77@users.noreply.github.com>
Co-authored-by: Claude Opus 4.8 <noreply@anthropic.com>
Co-authored-by: Raxxoor <manker_lol@hotmail.com>
@diegosouzapw
diegosouzapw deleted the fix/3419-opencode-deprecate branch June 11, 2026 15:33
oyi77 pushed a commit to oyi77/OmniRoute that referenced this pull request Jun 12, 2026
@diegosouzapw @oyi77
docs(opencode-provider): soft-deprecate in favor of @omniroute/openco…
4cb5e1e 
…de-plugin (cheaper-inference#3419) (cheaper-inference#3613)

Integrated into release/v3.8.21 (cheaper-inference#3419)
oyi77 pushed a commit to oyi77/OmniRoute that referenced this pull request Jun 12, 2026
@diegosouzapw @oyi77
docs(opencode-provider): soft-deprecate in favor of @omniroute/openco…
cc24d53 
…de-plugin (cheaper-inference#3419) (cheaper-inference#3613)

Integrated into release/v3.8.21 (cheaper-inference#3419)
oyi77 pushed a commit to oyi77/OmniRoute that referenced this pull request Jun 12, 2026
@diegosouzapw @oyi77
docs(opencode-provider): soft-deprecate in favor of @omniroute/openco…
27c9ced 
…de-plugin (cheaper-inference#3419) (cheaper-inference#3613)

Integrated into release/v3.8.21 (cheaper-inference#3419)
oyi77 pushed a commit to oyi77/OmniRoute that referenced this pull request Jun 12, 2026
@diegosouzapw @oyi77
docs(opencode-provider): soft-deprecate in favor of @omniroute/openco…
125d9e2 
…de-plugin (cheaper-inference#3419) (cheaper-inference#3613)

Integrated into release/v3.8.21 (cheaper-inference#3419)
oyi77 pushed a commit to oyi77/OmniRoute that referenced this pull request Jun 12, 2026
@diegosouzapw @oyi77
docs(opencode-provider): soft-deprecate in favor of @omniroute/openco…
c28e2f4 
…de-plugin (cheaper-inference#3419) (cheaper-inference#3613)

Integrated into release/v3.8.21 (cheaper-inference#3419)
oyi77 pushed a commit to oyi77/OmniRoute that referenced this pull request Jun 12, 2026
@diegosouzapw @oyi77
docs(opencode-provider): soft-deprecate in favor of @omniroute/openco…
6a28ca0 
…de-plugin (cheaper-inference#3419) (cheaper-inference#3613)

Integrated into release/v3.8.21 (cheaper-inference#3419)
oyi77 pushed a commit to oyi77/OmniRoute that referenced this pull request Jun 12, 2026
@diegosouzapw @oyi77
docs(opencode-provider): soft-deprecate in favor of @omniroute/openco…
ff5996e 
…de-plugin (cheaper-inference#3419) (cheaper-inference#3613)

Integrated into release/v3.8.21 (cheaper-inference#3419)
HouMinXi pushed a commit to HouMinXi/OmniRoute that referenced this pull request Aug 2, 2026
@diegosouzapw @herjarsa
@oyi77 @claude @dhaern
Release v3.8.21 (cheaper-inference#3593)
737dc4f 
* chore(release): open v3.8.21 development cycle

* fix: pass through valid max_tokens-truncated responses instead of fake 502 (cheaper-inference#3572) (cheaper-inference#3595)

* fix: /v1/completions returns legacy text-completion format, not chat (cheaper-inference#3571) (cheaper-inference#3596)

* fix: z.ai/GLM coding plan no longer shows Monthly 0% when no monthly cap (cheaper-inference#3580) (cheaper-inference#3597)

* docs: mark DISCOVERY_TOOL_DESIGN endpoints as Phase-2 not-yet-implemented (cheaper-inference#3498) (cheaper-inference#3599)

* fix(agent-bridge): add validate-only upstream-ca/test route (cheaper-inference#3488) (cheaper-inference#3600)

* fix(gamification): add level/badges/badges-earned profile routes (cheaper-inference#3484)

* security(oauth): migrate 5 public client_ids to resolvePublicCred (cheaper-inference#3493)

* fix(mcp): ship MCP server source closure in npm files + coverage gate (cheaper-inference#3578)

* fix: add reasoning token buffer for combo routing (fixes cheaper-inference#3587) (cheaper-inference#3588)

Integrated into release/v3.8.21

* Refactor: Extract chatCore phases into modular files (cheaper-inference#3598)

Integrated into release/v3.8.21 — chatCore phase modularization. Adjusted: re-derive idempotencyKey for the save path after the check moved into the module (co-authored). Thanks @oyi77!

* docs(changelog): credit cheaper-inference#3598 (chatCore modularization) + cheaper-inference#3588 (combo reasoning buffer)

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>

* fix(api): implement GET /api/guardrails + POST /api/guardrails/test, drop shadow/guardrails doc-fiction (cheaper-inference#3496) (cheaper-inference#3602)

Integrated into release/v3.8.21 — implements GET /api/guardrails + POST /api/guardrails/test, removes shadow/guardrails doc-fiction. TDD-validated (5/5) + check-docs-symbols/typecheck/eslint green.

* fix(gemini): isolate textual reasoning wrappers (cheaper-inference#3605)

Split-out PR C from cheaper-inference#3584. Isolates textual reasoning wrappers (<think>/<thinking>/<thought>/<internal_thought>, including malformed/open tags) into reasoning_content across both the non-streaming sanitizer and the Gemini streaming translator, with split-chunk buffering. Additive to the existing textual tool-call pipeline; does not touch the cheaper-inference#3569 native functionResponse path. Integrated into release/v3.8.21. Thanks @dhaern!

* fix(antigravity): normalize Gemini 3.5 Flash tier IDs (cheaper-inference#3603)

Split-out PR A from cheaper-inference#3584. Normalizes the Antigravity/agy Gemini 3.5 Flash tier IDs to clean public names (gemini-3.5-flash-low/medium/high), maps them to the live upstream IDs at the executor boundary, and removes Antigravity from the global model resolver so the executor owns wire normalization. Maintainer follow-up: kept gemini-3.5-flash-preview as a hidden backward-compat alias routing to the High tier (so saved combos/configs keep working). Live-validated the tier set via the agy CLI catalog. Integrated into release/v3.8.21. Thanks @dhaern!

* fix(agent-bridge): surface real MITM startup-failure cause, not always port 443 (cheaper-inference#3606) (cheaper-inference#3608)

Integrated into release/v3.8.21 (cheaper-inference#3606)

* fix(oauth): surface real Kiro import-token failure cause, not a bare 500 (cheaper-inference#3589) (cheaper-inference#3609)

Integrated into release/v3.8.21 (cheaper-inference#3589)

* docs(opencode-provider): soft-deprecate in favor of @omniroute/opencode-plugin (cheaper-inference#3419) (cheaper-inference#3613)

Integrated into release/v3.8.21 (cheaper-inference#3419)

* fix(usage): normalize Antigravity and agy provider quotas (cheaper-inference#3604)

Split-out PR B from cheaper-inference#3584. Normalizes Antigravity/agy provider quotas: prefers retrieveUserQuota for live consumption, falls back to fetchAvailableModels and local usage_history, sanitizes cached Provider Limits so retired upstream IDs are not re-exposed, and schedules a deduplicated post-usage refresh. Maintainer follow-up: decoupled the post-usage refresh via a lightweight usageEvents bus (usageHistory no longer dynamic-imports providerLimits) so it does not pull the executors/translator graph into the typecheck-core surface — typecheck:core stays at 0. Integrated into release/v3.8.21. Thanks @dhaern!

* feat(cli): add autostart on/off/toggle shorthand for headless serve mode (cheaper-inference#3331) (cheaper-inference#3614)

Integrated into release/v3.8.21 (cheaper-inference#3331)

* docs(changelog): credit cheaper-inference#3603 (Flash tier IDs) + cheaper-inference#3604 (provider quotas) + cheaper-inference#3605 (reasoning wrappers)

Co-authored-by: diegosouzapw <diegosouza.pw@gmail.com>

* fix(review): resolve findings from /review-reviews battery (v3.8.21 hardening) (cheaper-inference#3618)

Pre-release hardening from the /review-reviews battery — 15 findings resolved (L1-L13,L15) + L14 live-verified WONTFIX, convergence re-review clean. lint/typecheck:core/test:vitest(146)/build green; zero new test:unit failures vs baseline 6d24708.

* chore(release): v3.8.21 CHANGELOG + i18n + env-doc sync

---------

Co-authored-by: Hernan Javier Ardila Sanchez <hjasgr@gmail.com>
Co-authored-by: Paijo <14921983+oyi77@users.noreply.github.com>
Co-authored-by: Claude Opus 4.8 <noreply@anthropic.com>
Co-authored-by: Raxxoor <manker_lol@hotmail.com>
Poid-ZA pushed a commit to Poid-ZA/OmniRoute that referenced this pull request Aug 5, 2026
@diegosouzapw @herjarsa
@oyi77 @claude @dhaern
Release v3.8.21 (cheaper-inference#3593)
a7d759e 
* chore(release): open v3.8.21 development cycle

* fix: pass through valid max_tokens-truncated responses instead of fake 502 (cheaper-inference#3572) (cheaper-inference#3595)

* fix: /v1/completions returns legacy text-completion format, not chat (cheaper-inference#3571) (cheaper-inference#3596)

* fix: z.ai/GLM coding plan no longer shows Monthly 0% when no monthly cap (cheaper-inference#3580) (cheaper-inference#3597)

* docs: mark DISCOVERY_TOOL_DESIGN endpoints as Phase-2 not-yet-implemented (cheaper-inference#3498) (cheaper-inference#3599)

* fix(agent-bridge): add validate-only upstream-ca/test route (cheaper-inference#3488) (cheaper-inference#3600)

* fix(gamification): add level/badges/badges-earned profile routes (cheaper-inference#3484)

* security(oauth): migrate 5 public client_ids to resolvePublicCred (cheaper-inference#3493)

* fix(mcp): ship MCP server source closure in npm files + coverage gate (cheaper-inference#3578)

* fix: add reasoning token buffer for combo routing (fixes cheaper-inference#3587) (cheaper-inference#3588)

Integrated into release/v3.8.21

* Refactor: Extract chatCore phases into modular files (cheaper-inference#3598)

Integrated into release/v3.8.21 — chatCore phase modularization. Adjusted: re-derive idempotencyKey for the save path after the check moved into the module (co-authored). Thanks @oyi77!

* docs(changelog): credit cheaper-inference#3598 (chatCore modularization) + cheaper-inference#3588 (combo reasoning buffer)

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>

* fix(api): implement GET /api/guardrails + POST /api/guardrails/test, drop shadow/guardrails doc-fiction (cheaper-inference#3496) (cheaper-inference#3602)

Integrated into release/v3.8.21 — implements GET /api/guardrails + POST /api/guardrails/test, removes shadow/guardrails doc-fiction. TDD-validated (5/5) + check-docs-symbols/typecheck/eslint green.

* fix(gemini): isolate textual reasoning wrappers (cheaper-inference#3605)

Split-out PR C from cheaper-inference#3584. Isolates textual reasoning wrappers (<think>/<thinking>/<thought>/<internal_thought>, including malformed/open tags) into reasoning_content across both the non-streaming sanitizer and the Gemini streaming translator, with split-chunk buffering. Additive to the existing textual tool-call pipeline; does not touch the cheaper-inference#3569 native functionResponse path. Integrated into release/v3.8.21. Thanks @dhaern!

* fix(antigravity): normalize Gemini 3.5 Flash tier IDs (cheaper-inference#3603)

Split-out PR A from cheaper-inference#3584. Normalizes the Antigravity/agy Gemini 3.5 Flash tier IDs to clean public names (gemini-3.5-flash-low/medium/high), maps them to the live upstream IDs at the executor boundary, and removes Antigravity from the global model resolver so the executor owns wire normalization. Maintainer follow-up: kept gemini-3.5-flash-preview as a hidden backward-compat alias routing to the High tier (so saved combos/configs keep working). Live-validated the tier set via the agy CLI catalog. Integrated into release/v3.8.21. Thanks @dhaern!

* fix(agent-bridge): surface real MITM startup-failure cause, not always port 443 (cheaper-inference#3606) (cheaper-inference#3608)

Integrated into release/v3.8.21 (cheaper-inference#3606)

* fix(oauth): surface real Kiro import-token failure cause, not a bare 500 (cheaper-inference#3589) (cheaper-inference#3609)

Integrated into release/v3.8.21 (cheaper-inference#3589)

* docs(opencode-provider): soft-deprecate in favor of @omniroute/opencode-plugin (cheaper-inference#3419) (cheaper-inference#3613)

Integrated into release/v3.8.21 (cheaper-inference#3419)

* fix(usage): normalize Antigravity and agy provider quotas (cheaper-inference#3604)

Split-out PR B from cheaper-inference#3584. Normalizes Antigravity/agy provider quotas: prefers retrieveUserQuota for live consumption, falls back to fetchAvailableModels and local usage_history, sanitizes cached Provider Limits so retired upstream IDs are not re-exposed, and schedules a deduplicated post-usage refresh. Maintainer follow-up: decoupled the post-usage refresh via a lightweight usageEvents bus (usageHistory no longer dynamic-imports providerLimits) so it does not pull the executors/translator graph into the typecheck-core surface — typecheck:core stays at 0. Integrated into release/v3.8.21. Thanks @dhaern!

* feat(cli): add autostart on/off/toggle shorthand for headless serve mode (cheaper-inference#3331) (cheaper-inference#3614)

Integrated into release/v3.8.21 (cheaper-inference#3331)

* docs(changelog): credit cheaper-inference#3603 (Flash tier IDs) + cheaper-inference#3604 (provider quotas) + cheaper-inference#3605 (reasoning wrappers)

Co-authored-by: diegosouzapw <diegosouza.pw@gmail.com>

* fix(review): resolve findings from /review-reviews battery (v3.8.21 hardening) (cheaper-inference#3618)

Pre-release hardening from the /review-reviews battery — 15 findings resolved (L1-L13,L15) + L14 live-verified WONTFIX, convergence re-review clean. lint/typecheck:core/test:vitest(146)/build green; zero new test:unit failures vs baseline 797de43.

* chore(release): v3.8.21 CHANGELOG + i18n + env-doc sync

---------

Co-authored-by: Hernan Javier Ardila Sanchez <hjasgr@gmail.com>
Co-authored-by: Paijo <14921983+oyi77@users.noreply.github.com>
Co-authored-by: Claude Opus 4.8 <noreply@anthropic.com>
Co-authored-by: Raxxoor <manker_lol@hotmail.com>
tkgo11 pushed a commit to tkgo11/OmniRoute that referenced this pull request Sep 23, 2026
@diegosouzapw
docs(opencode-provider): soft-deprecate in favor of @omniroute/openco…
fb3e692 
…de-plugin (cheaper-inference#3419) (cheaper-inference#3613)

Integrated into release/v3.8.21 (cheaper-inference#3419)
muhamadgalihsaputra pushed a commit to niyatna/NiyatnaRoute that referenced this pull request Sep 27, 2026
@diegosouzapw @herjarsa
@oyi77 @claude @dhaern
Release v3.8.21 (cheaper-inference#3593)
cab8cde 
* chore(release): open v3.8.21 development cycle

* fix: pass through valid max_tokens-truncated responses instead of fake 502 (cheaper-inference#3572) (cheaper-inference#3595)

* fix: /v1/completions returns legacy text-completion format, not chat (cheaper-inference#3571) (cheaper-inference#3596)

* fix: z.ai/GLM coding plan no longer shows Monthly 0% when no monthly cap (cheaper-inference#3580) (cheaper-inference#3597)

* docs: mark DISCOVERY_TOOL_DESIGN endpoints as Phase-2 not-yet-implemented (cheaper-inference#3498) (cheaper-inference#3599)

* fix(agent-bridge): add validate-only upstream-ca/test route (cheaper-inference#3488) (cheaper-inference#3600)

* fix(gamification): add level/badges/badges-earned profile routes (cheaper-inference#3484)

* security(oauth): migrate 5 public client_ids to resolvePublicCred (cheaper-inference#3493)

* fix(mcp): ship MCP server source closure in npm files + coverage gate (cheaper-inference#3578)

* fix: add reasoning token buffer for combo routing (fixes cheaper-inference#3587) (cheaper-inference#3588)

Integrated into release/v3.8.21

* Refactor: Extract chatCore phases into modular files (cheaper-inference#3598)

Integrated into release/v3.8.21 — chatCore phase modularization. Adjusted: re-derive idempotencyKey for the save path after the check moved into the module (co-authored). Thanks @oyi77!

* docs(changelog): credit cheaper-inference#3598 (chatCore modularization) + cheaper-inference#3588 (combo reasoning buffer)

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>

* fix(api): implement GET /api/guardrails + POST /api/guardrails/test, drop shadow/guardrails doc-fiction (cheaper-inference#3496) (cheaper-inference#3602)

Integrated into release/v3.8.21 — implements GET /api/guardrails + POST /api/guardrails/test, removes shadow/guardrails doc-fiction. TDD-validated (5/5) + check-docs-symbols/typecheck/eslint green.

* fix(gemini): isolate textual reasoning wrappers (cheaper-inference#3605)

Split-out PR C from cheaper-inference#3584. Isolates textual reasoning wrappers (<think>/<thinking>/<thought>/<internal_thought>, including malformed/open tags) into reasoning_content across both the non-streaming sanitizer and the Gemini streaming translator, with split-chunk buffering. Additive to the existing textual tool-call pipeline; does not touch the cheaper-inference#3569 native functionResponse path. Integrated into release/v3.8.21. Thanks @dhaern!

* fix(antigravity): normalize Gemini 3.5 Flash tier IDs (cheaper-inference#3603)

Split-out PR A from cheaper-inference#3584. Normalizes the Antigravity/agy Gemini 3.5 Flash tier IDs to clean public names (gemini-3.5-flash-low/medium/high), maps them to the live upstream IDs at the executor boundary, and removes Antigravity from the global model resolver so the executor owns wire normalization. Maintainer follow-up: kept gemini-3.5-flash-preview as a hidden backward-compat alias routing to the High tier (so saved combos/configs keep working). Live-validated the tier set via the agy CLI catalog. Integrated into release/v3.8.21. Thanks @dhaern!

* fix(agent-bridge): surface real MITM startup-failure cause, not always port 443 (cheaper-inference#3606) (cheaper-inference#3608)

Integrated into release/v3.8.21 (cheaper-inference#3606)

* fix(oauth): surface real Kiro import-token failure cause, not a bare 500 (cheaper-inference#3589) (cheaper-inference#3609)

Integrated into release/v3.8.21 (cheaper-inference#3589)

* docs(opencode-provider): soft-deprecate in favor of @omniroute/opencode-plugin (cheaper-inference#3419) (cheaper-inference#3613)

Integrated into release/v3.8.21 (cheaper-inference#3419)

* fix(usage): normalize Antigravity and agy provider quotas (cheaper-inference#3604)

Split-out PR B from cheaper-inference#3584. Normalizes Antigravity/agy provider quotas: prefers retrieveUserQuota for live consumption, falls back to fetchAvailableModels and local usage_history, sanitizes cached Provider Limits so retired upstream IDs are not re-exposed, and schedules a deduplicated post-usage refresh. Maintainer follow-up: decoupled the post-usage refresh via a lightweight usageEvents bus (usageHistory no longer dynamic-imports providerLimits) so it does not pull the executors/translator graph into the typecheck-core surface — typecheck:core stays at 0. Integrated into release/v3.8.21. Thanks @dhaern!

* feat(cli): add autostart on/off/toggle shorthand for headless serve mode (cheaper-inference#3331) (cheaper-inference#3614)

Integrated into release/v3.8.21 (cheaper-inference#3331)

* docs(changelog): credit cheaper-inference#3603 (Flash tier IDs) + cheaper-inference#3604 (provider quotas) + cheaper-inference#3605 (reasoning wrappers)

Co-authored-by: diegosouzapw <diegosouza.pw@gmail.com>

* fix(review): resolve findings from /review-reviews battery (v3.8.21 hardening) (cheaper-inference#3618)

Pre-release hardening from the /review-reviews battery — 15 findings resolved (L1-L13,L15) + L14 live-verified WONTFIX, convergence re-review clean. lint/typecheck:core/test:vitest(146)/build green; zero new test:unit failures vs baseline 408d91a2c.

* chore(release): v3.8.21 CHANGELOG + i18n + env-doc sync

---------

Co-authored-by: Hernan Javier Ardila Sanchez <hjasgr@gmail.com>
Co-authored-by: Paijo <14921983+oyi77@users.noreply.github.com>
Co-authored-by: Claude Opus 4.8 <noreply@anthropic.com>
Co-authored-by: Raxxoor <manker_lol@hotmail.com>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Reviewers

No reviews

Assignees

No one assigned

Labels

None yet

Projects

None yet

Milestone

No milestone

Development

Successfully merging this pull request may close these issues.

1 participant

@diegosouzapw