@bivelio/savings-layer 0.5.0-rc.1 → 0.5.0

package/README.md

@@ -170,6 +170,9 @@ and the free **Measure** license, which runs the layer in shadow mode on your ow
 traffic and shows what it would have saved, valued with your provider's counter. The verifier that replays the optimizer on your own data
 lives in your dashboard once you sign in.
 
+**Upgrading from 0.4?** Every breaking change of 0.5, with a before and after for
+each, is in [Migrating from 0.4 to 0.5](https://savings.bivelio.com/en/migrate-to-0.5).
+
 ## Set it up in one command
 
 ```bash
@@ -914,7 +917,7 @@ Ask for a response contract and you also get `data`, validated against your sche
 ```ts
 const { data, text } = await layer.generate<{ answer: number }>({
   model: "openai/gpt-4o-mini",
-  messages: [{ role: "user", content: "What is 2 + 2 * 3?" }],
+  messages: [{ role: "user", content: "How many legs does a spider have?" }],
   response: {
     contractId: "answer:v1",
     schema: {
@@ -930,6 +933,11 @@ const { data, text } = await layer.generate<{ answer: number }>({
 console.log(data.answer); // 8
 ```
 
+A question the layer answers on its own (the arithmetic of the first example)
+never reaches the model, so its `data` is the resolver's `{ value: 8 }`, not
+your schema's shape: `report.qualityLedger` says the contract was not met.
+Ask the model for anything you want in your contract's shape.
+
 </details>
 
 <details>

package/dist/gateway/bin.js

@@ -79,7 +79,7 @@ td.test(n.betas))return null;let r=ct(e,t);if(r===null||nd(r))return null;let o=
 let r=Object.entries(ks).map(([o,i])=>`${o}\u2192${i}`).join(", ");return e!=="1"?{linea:"sibling model routing: OFF (BVSALA_GATEWAY_SIBLING_MODEL=1 turns it on)",aviso:false}:n?{linea:`sibling model r\
 outing: ARMED (${r}; ON if your licence includes sibling_model_routing; fail-closed). Changes the model you see in the response; cache reads cost 0.025\xD7 instead of 0.1\xD7 (list price). Turning it on mid\
 -conversation pays one cold prefix write, once. If the provider rejects the rewritten request (400), it is re-sent unchanged once`,aviso:false}:{linea:"sibling model routing: NOT ARMED - BVSALA_GATEWAY_E\
-LISION=0 turns off the levers",aviso:true}}var xs=new Set;function ee(e,t){xs.has(e)||(xs.add(e),console.warn(t));}var ye="0.5.0-rc.1",fo="x-bvsala-version";var Ut=class{options;now;flushIntervalMs;maxBatch;maxBuffer;buffer=[];seatId;ci;timer;inflight=null;fallosSeguidos=0;constructor(t){this.options=t,this.now=t.now??(()=>Date.now()),this.flushIntervalMs=
+LISION=0 turns off the levers",aviso:true}}var xs=new Set;function ee(e,t){xs.has(e)||(xs.add(e),console.warn(t));}var ye="0.5.0",fo="x-bvsala-version";var Ut=class{options;now;flushIntervalMs;maxBatch;maxBuffer;buffer=[];seatId;ci;timer;inflight=null;fallosSeguidos=0;constructor(t){this.options=t,this.now=t.now??(()=>Date.now()),this.flushIntervalMs=
 t.flushIntervalMs??15e3,this.maxBatch=t.maxBatch??50,this.maxBuffer=t.maxBuffer??1e3,this.ci=t.ci??false;}setSeat(t,n){t&&(this.seatId=t),this.ci=n;}record(t){let n=this.now();this.buffer.push({report:t,at:n});
 let r=rd(t);r&&this.buffer.push({report:r,at:n}),this.buffer.length>this.maxBuffer&&this.buffer.splice(0,this.buffer.length-this.maxBuffer),this.buffer.length>=this.maxBatch&&this.flush();}start(){this.
 timer||(this.timer=setInterval(()=>{this.flush();},this.flushIntervalMs),this.timer.unref?.());}stop(){this.timer&&(clearInterval(this.timer),this.timer=void 0),this.flush();}async stopAndDrain(t){this.timer&&

package/dist/index.d.ts

@@ -154,7 +154,7 @@ declare function precioDeLista(model: string, at?: Date | number): CostModel | u
  * on the customer's machine. `test/version-declarada.test.ts` ties this
  * constant to package.json so the two can never drift.
  */
-declare const BV_SALA_VERSION = "0.5.0-rc.1";
+declare const BV_SALA_VERSION = "0.5.0";
 
 /** How the layer served a request that is eligible for a verification check. */
 type VerificationStratum = 

package/dist/index.js

@@ -28,7 +28,7 @@ for(let n of Yp[t]??[])if(!(n.vigenteDesde!==void 0&&e<n.vigenteDesde)&&!(n.vige
 p-research","o4-mini-deep-research","gpt-5.5-pro","gpt-5.4-pro","gpt-5.2-pro","gpt-5-pro","gpt-5.5-cyber","gemini-3.8-flash-tts","gemini-3.8-flash-lite-tts","gemini-3.1-flash-lite-image"],Xo={...qt,...Object.
 fromEntries(Qp.map(t=>[t,null]))},Xp=ti(Xo);function gn(t,e){let n=Ut(t);return n===void 0?void 0:Ka(n,Ja(e))}var ef=Object.fromEntries(Object.keys(Xo).map(t=>[t,Xo[t]===null?null:t]));function Ut(t){
 return ei(ef,Xp,t)??void 0}var tf=/^\d{4}-?\d{2}-?\d{2}$/;function st(t,e){let n=Ut(t);if(n===void 0)return;let r=t.lastIndexOf("/"),o=r>=0?t.slice(r+1):t;if(!(n!==t&&n!==o&&!tf.test(o.slice(n.length+
-1))))return Ka(n,Ja(e))}var He="0.5.0-rc.1",fr="x-bvsala-version";var nf=["exact","task_equivalent","not_verified","bounded_degradation","rejected"];function rf(t){return {v:1,task:t.task,goal:t.goal,rules:t.rules??[],facts:t.facts??[],refs:t.refs??[],...t.delta?{delta:t.delta}:{},outputContract:t.outputContract}}var mr=class{id;exact=false;constructor(e="heuristic"){this.id=e;}count(e){if(e.length===0)return 0;let n=e.match(/[A-Za-z]+|\d+|\s+|[^\sA-Za-z\d]/g)??[],r=0;for(let o of n){let i=o[0];if(/\s/.test(i)){let s=(o.
+1))))return Ka(n,Ja(e))}var He="0.5.0",fr="x-bvsala-version";var nf=["exact","task_equivalent","not_verified","bounded_degradation","rejected"];function rf(t){return {v:1,task:t.task,goal:t.goal,rules:t.rules??[],facts:t.facts??[],refs:t.refs??[],...t.delta?{delta:t.delta}:{},outputContract:t.outputContract}}var mr=class{id;exact=false;constructor(e="heuristic"){this.id=e;}count(e){if(e.length===0)return 0;let n=e.match(/[A-Za-z]+|\d+|\s+|[^\sA-Za-z\d]/g)??[],r=0;for(let o of n){let i=o[0];if(/\s/.test(i)){let s=(o.
 match(/\n/g)??[]).length;r+=s+Math.floor(Math.max(0,o.length-1)/2);}else /[A-Za-z]/.test(i)?r+=Math.max(1,Math.ceil(o.length/4)):/\d/.test(i)?r+=Math.max(1,Math.ceil(o.length/2)):r+=1;}return Math.max(r,
 Math.ceil(e.length/4))}},St=new mr;var hr=class{estimators=new Map;modelToTokenizer=new Map;fallbackId;constructor(e=St){this.fallbackId=e.id,this.estimators.set(e.id,e);}register(e){return this.estimators.set(e.id,e),this}mapModel(e,n){
 return this.modelToTokenizer.set(e,n),this}resolve(e){let n=e.tokenizer?this.estimators.get(e.tokenizer):void 0;if(n)return n;if(e.model){let r=this.modelToTokenizer.get(e.model);if(r){let o=this.estimators.

package/docs/terminales/README.md

@@ -302,8 +302,8 @@ would fail it.
 | Lever | Applies to | What changes | What does not |
 |---|---|---|---|
 | Cache TTL (`cache_ttl_tuning`, on by default with PRO) | `/v1/messages` requests where a large part of the prompt is new (session start, compaction); a chained turn travels untouched | every `cache_control` with `ttl: "1h"` becomes `"5m"`, the top-level one included (a 1-hour top-level marker behind 5-minute block markers would be a 400) | a key named `cache_control` inside a tool call's `input` or a tool's `input_schema` is content and stays as it is; so does the advisor's `caching`; headers are untouched |
-| Per-turn effort (`BVSALA_GATEWAY_ESFUERZO=low` or `medium`) | `/v1/messages` for `claude-fable-5-1`, `claude-mythos-5-1`, `claude-opus-5-5` and `claude-opus-5`, when the level is below the model's default effort (`high`; `medium` on `claude-opus-5-5`) and the request carries no `output_config` of its own (top level or in a message); a level at or above the default leaves the request untouched | one message appended at the end of `messages`, `{"role":"system","content":[],"output_config":{"effort":…}}`, and `mid-conversation-output-config-2026-07-01` appended after your `anthropic-beta` value | your `anthropic-beta` value is kept verbatim in front; nothing earlier in `messages` moves |
-| Sibling model (`BVSALA_GATEWAY_MODELO_HERMANO=1`) | `/v1/messages` and `/v1/messages/count_tokens` for `claude-fable-5`, except with a forced `tool_choice` or the 1M-context beta | `model` becomes `claude-fable-5-1` | headers are untouched |
+| Per-turn effort (`BVSALA_GATEWAY_EFFORT=low` or `medium`) | `/v1/messages` for `claude-fable-5-1`, `claude-mythos-5-1`, `claude-opus-5-5` and `claude-opus-5`, when the level is below the model's default effort (`high`; `medium` on `claude-opus-5-5`) and the request carries no `output_config` of its own (top level or in a message); a level at or above the default leaves the request untouched | one message appended at the end of `messages`, `{"role":"system","content":[],"output_config":{"effort":…}}`, and `mid-conversation-output-config-2026-07-01` appended after your `anthropic-beta` value | your `anthropic-beta` value is kept verbatim in front; nothing earlier in `messages` moves |
+| Sibling model (`BVSALA_GATEWAY_SIBLING_MODEL=1`) | `/v1/messages` and `/v1/messages/count_tokens` for `claude-fable-5`, except with a forced `tool_choice` or the 1M-context beta | `model` becomes `claude-fable-5-1` | headers are untouched |
 | Context-editing dials (`BVSALA_GATEWAY_DIALES_ELISION=1`; measured as a net loss, off) | `/v1/messages` | your `clear_tool_uses_20250919` edit in `context_management` is replaced by ours | your other edits are kept, `clear_thinking_20251015` first; your `anthropic-beta` value is kept verbatim, with `context-management-2025-06-27` appended if it was not there |
 | Usage in Chat Completions streams (`BVSALA_GATEWAY_USAGE_INJECT=1`, off) | streaming `…/chat/completions` requests without `stream_options` of their own | `"stream_options":{"include_usage":true},` inserted as text right after the opening `{` | the rest of the body, byte for byte (nothing is re-serialized); headers are untouched; a 400 or 422 gets your original request once more |
 

package/docs/terminales/codex.md

@@ -47,7 +47,7 @@ codex -c openai_base_url=http://127.0.0.1:8402/v1                 # ventana 2
 
 Codex añade `/responses` (`POST {base_url}/responses`, siempre `stream: true`)
 [V]. Si en el 8402 ya tienes el gateway de Anthropic para otros terminales,
-arranca este en otro puerto (`--puerto 8403`) y apunta ahí: un gateway habla con
+arranca este en otro puerto (`--port 8403`) y apunta ahí: un gateway habla con
 un solo upstream.
 
 No lo escribas en `~/.codex/config.toml`: ahí se quedaría fijo y Codex fallaría

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@bivelio/savings-layer",
-  "version": "0.5.0-rc.1",
+  "version": "0.5.0",
   "description": "The savings layer for your LLM bill: a library inside your code that lets through only what has to be paid for and returns a SavingsReport per call, measured with your provider's own meter and kept apart from estimates. Commercial license, paid per seat, commission only on measured savings. Your prompts never leave your infrastructure.",
   "keywords": [
     "llm",