[ggiak]

Summary

Wires `claude-fable-5` across the full model pipeline so it is available as a first-class model in OmniRoute.

• Provider registry (`open-sse/config/providerRegistry.ts`) — adds `claude-fable-5` to the `cc` (Claude Code OAuth) and `kiro` model lists with 1M context / 128k output tokens
• Pricing (`src/shared/constants/pricing.ts`) — adds `CLAUDE_FABLE_5_PRICING` constant ($15/$75 per 1M tokens, same tier as Opus 4.8); wired into `cc`, `kiro`, and the shared `anthropic` blocks
• Model spec (`src/shared/constants/modelSpecs.ts`) — adds spec entry: 1M context, 128k output, thinking support (same adaptive-thinking constraints as Opus 4.8)
• Static models (`src/lib/providers/staticModels.ts`) — adds to `claude` provider list
• Fast mode (`src/lib/providers/claudeFastMode.ts`) — adds to default fast-mode model list
• Settings default (`src/lib/db/settings.ts`) — adds to default `claudeFastMode.supportedModels`
• 1M context (`open-sse/services/claudeCodeCompatible.ts`) — adds to `CONTEXT_1M_SUPPORTED_MODELS`
• Fallback chain (`open-sse/services/modelFamilyFallback.ts`) — `claude-fable-5 → claude-opus-4-8 → claude-opus-4-7 → claude-sonnet-4-6`
• Cost data (`open-sse/services/providerCostData.ts`) — adds cost entry at $15/$75 per 1M
• CHANGELOG — added entry under `[Unreleased]`

Test plan

• New test in `tests/unit/catalog-updates-v3x.test.ts` — verifies `claude-fable-5` appears in both `cc` and `kiro` catalogs with correct `contextLength`, `maxOutputTokens`, and pricing entries in both channels
• Targeted test passes: `npx tsx --test tests/unit/catalog-updates-v3x.test.ts` → 5/5 pass
• `npm run lint` passes
• Full `npm run test:unit` — pre-existing failures unrelated to this PR:
  • `kiro-multi-account-isolation` — assertion flake present on `main` before this commit
  • Several `claude-*` tests fail due to missing `uuid` package in dev environment (also pre-existing on `main`)
  • No new failures introduced by this PR

[gemini-code-assist]

Code Review

This pull request integrates the new claude-fable-5 model across the application, updating provider registries, pricing structures, model specifications, fast mode configurations, and fallback definitions, as well as adding unit tests. The review feedback points out two important issues: a notation mismatch in model family fallbacks that could cause fallback failures for dot-notation providers, and a missing Bedrock alias in the model specifications that would cause Bedrock-routed requests to fall back to default token limits.

Important

The consumer version of Gemini Code Assist on GitHub is being sunset. Starting June 18, 2026, new organization installations will be blocked, and all code review activity will officially cease on July 17, 2026.
For more details on the timeline and next steps, please review the Help Documentation.

[ggiak]

Addressed both review comments in 50af497:

[High] Notation mismatch in modelFamilyFallback.ts

• getNextFamilyFallback now normalizes dots→hyphens on the lookup key so kiro/claude-opus-4.8 correctly finds its MODEL_FAMILIES entry
• Candidates are resolved against the provider's registered model IDs via getRegistryEntry — dot-notation providers (kiro, github) receive dot-notation fallbacks; hyphen-notation providers (cc, vertex) receive hyphen-notation. No more mismatched IDs being passed to the executor.
• claude-fable-5 moved out of the Claude Opus family into its own Mythos family with fallback chain: claude-fable-5 → claude-sonnet-4-6 → claude-opus-4-8

[Medium] Missing Bedrock aliases in modelSpecs.ts

• Added BEDROCK_CLAUDE_ALIASES("claude-fable-5") to the spec entry — Bedrock-routed requests now resolve the correct 128k max output tokens instead of falling back to the 8k default.

[diegosouzapw]

Thanks @ggiak! Wiring claude-fable-5 across the full pipeline (cc+kiro registries, model spec, pricing, fast mode, 1M-context header, fallback family) is exactly what a new flagship needs, and the catalog test covers the registration. I pushed two adjustments: (1) reordered the fallback chain from [sonnet-4-6, opus-4-8] to [opus-4-8, opus-4-7, sonnet-4-6] — a flagship should fall to the next-best Opus before the cheaper Sonnet (and it now matches your CHANGELOG line); (2) added tests for getNextFamilyFallback's new dot-vs-hyphen notation resolver (the hunk that affects all families) — kiro/claude-fable-5 → kiro/claude-opus-4.8, plus the cc→claude alias normalization. Typecheck clean, tests green. Shipping in the next release. 🙌

[kilo-code-bot]

Kilo Code Review could not run — your account is out of credits.

Add credits or switch to a free model to enable reviews on this change.

[chirag127]

why we have duplication of the info like context lenght output token limit in many places