fix(pricing): price OpenRouter turns from the lake and declared rates

[Hmbown]

Fixes #6690

Cause

• The OpenRouter /v1/models refresh failed as a whole on any single bad row, so the OpenRouter lake scope stayed failed/invalid_response and fresh_provider_live_pricing_quote_at never returned a rate. Every OpenRouter turn showed "rate unavailable". The live list has two kinds of row that did this:
  • ~-prefixed "latest" aliases (~deepseek/deepseek-pro-latest, 18 in the live list), which fail valid_catalog_model_id.
  • six routers (openrouter/auto, auto-beta, fusion, pareto-code, bodybuilder, typesafe/jev-router) that publish "prompt":"-1","completion":"-1". parse_price rejected negatives and the conversion collected with ?.
• The main interactive turn (turn_loop dispatch boundary) froze its quote only from the lake, never from the operator-declared [[custom_models]] rate that effective_route_envelope (background/review) already honors.

Fix

• parse_openrouter_models_response still requires a {"data": [...]} list but decodes each row on its own. It drops ~ alias rows, and it skips and counts (with a tracing::warn!) rows that do not decode or whose id the catalog cannot hold. fetch_catalog_delta also skips and counts rows with an unreadable or implausible price. A structurally invalid response, or one where no row survives, is still InvalidResponse. The secret-in-id guard still rejects the whole list.
• A negative OpenRouter price means "no fixed rate": the row is kept with cost: None, never with a partial rate.
• New CodewhaleClient::configured_pricing_quote_at and client::main_turn_pricing_quote_at. The turn loop now prefers a declared rate, scoped to the installed client's endpoint fingerprint, before the lake quote.
• New provider_catalog_live::declared_or_catalog_quote, used at all three dispatch boundaries (main turn, effective_route_envelope, EffectiveRouteEnvelope::capture). A declared rate wins. A declaration with no rates yields to the same endpoint's catalog price, and stays frozen rate-less only when the catalog has none, so a same-named bundled price still cannot fill it.
• EffectiveRouteEnvelope::audit: a pinned openrouter_vendor still blocks the aggregate catalog price (RoutingDependentPrice), but no longer blocks an explicit declared (UserOverride) rate for the exact route.

Against the saved live list (458 rows), the refresh now succeeds with 440 rows, 434 of them priced. deepseek/deepseek-v4-pro prices at 0.95526 / 1.91052 USD per M (cache read 0.079605).

Still open: a server-tag-suffixed route the catalog does not list (the reporter's deepseek/deepseek-v4.1-flash:nitro) still needs an exact catalog row or a declared [[custom_models]] rate, which now prices it on the main turn. Tag prices can differ from the base id, so they are not stripped.

Tests

Each new or strengthened test fails with its fix reverted and passes with it. They assert the actual rates and cost, not just that a quote exists:

• client::tests::fetch_catalog_delta_skips_openrouter_latest_aliases_and_keeps_prices: live-shaped fixture (~ alias, -1 router, undecodable row, unreadable price, bad id, real deepseek/deepseek-v4-pro row). It also checks that structurally invalid bodies are still rejected.
• client::tests::provider_live_price_parsers_reject_present_bad_rates_but_keep_zero_and_omission: updated so a negative OpenRouter price gives cost: None.
• cost_status::tests::main_turn_rateless_declaration_yields_to_the_endpoint_catalog_price
• cost_status::tests::openrouter_vendor_pin_is_priced_by_an_explicit_declared_rate
• core::engine::tests::main_turn_dispatch_freezes_declared_custom_model_rate: now asserts the frozen rates.

fixed:    test result: ok. 10 passed; 0 failed; 0 ignored; 0 measured; 13793 filtered out
reverted (neg price, rate-less fallthrough, vendor pin):
          test result: FAILED. 0 passed; 3 failed; 0 ignored; 0 measured; 13800 filtered out
reverted (per-row decode skip):
          test result: FAILED. 0 passed; 1 failed; 0 ignored; 0 measured; 13802 filtered out
related (fetch_catalog openrouter configured_model effective_route provider_live_pricing
         exact_turn_snapshot main_turn cost_status provider_catalog_live):
          test result: ok. 120 passed; 0 failed; 0 ignored; 0 measured; 13683 filtered out

Only focused cargo test -p codewhale-tui --lib runs happened locally. The full suite is left to CI.

🤖 Generated with Claude Code

https://claude.ai/code/session_014ZwqatxgVFxHvovngywnks

[Copilot]

Copilot was unable to review this pull request because the user who requested the review has reached their quota limit.

[chatgpt-codex-connector]

Codex Review Summary

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

Review	Status	Commit	Review trigger
📝 Code Review	✅ Completed 2026-09-28T03:06:31.894112Z	2412a6e	PR opened

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.

[gitguardian]

️✅ There are no secrets present in this pull request anymore.

If these secrets were true positive and are still valid, we highly recommend you to revoke them.
While these secrets were previously flagged, we no longer have a reference to the
specific commits where they were detected. Once a secret has been leaked into a git
repository, you should consider it compromised, even if it was deleted immediately.
Find here more information about risks.

---

^{_{🦉 GitGuardian detects secrets in your source code to help developers and security teams secure the modern development process. You are seeing this because you or someone else with access to this repository has authorized GitGuardian to scan your pull request.}}