@suluk/models 0.1.2 → 0.1.3
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- package/package.json +1 -1
- package/src/types.ts +8 -2
package/package.json
CHANGED
|
@@ -1,6 +1,6 @@
|
|
|
1
1
|
{
|
|
2
2
|
"name": "@suluk/models",
|
|
3
|
-
"version": "0.1.
|
|
3
|
+
"version": "0.1.3",
|
|
4
4
|
"description": "A weekly, PUBLIC-DATA-ONLY catalog of OpenRouter models + a selector: a suluk skill declares NEEDS (hard filters) + a small PREFERENCE (a named profile), and selectModel picks the best CURRENT model — never a hard-coded id. Decidable OpenRouter facts as numbers; noisy benchmarks as COARSE TIERS with {source, asOf}; no cross-axis composite (blending is the selector's job). CANDIDATE tooling — NOT official OAS.",
|
|
5
5
|
"publishConfig": {
|
|
6
6
|
"access": "public"
|
package/src/types.ts
CHANGED
|
@@ -1,6 +1,12 @@
|
|
|
1
1
|
/**
|
|
2
|
-
* @suluk/models — the catalog schema (council wf_729cde52-cc7). A row is
|
|
3
|
-
*
|
|
2
|
+
* @suluk/models — the catalog schema (council wf_729cde52-cc7). A row is keyed BY MODEL (id, capabilities, benchmark
|
|
3
|
+
* tiers, context window — all per-model). Per-ENDPOINT axes (price/region/data-retention/latency, which differ across
|
|
4
|
+
* the provider endpoints one model fans out to) belong in a future optional `endpoints[]` sub-list (keying micro-panel
|
|
5
|
+
* wf_27de1bec-a42: model-keyed HYBRID, @0.6 — RESERVED, not built: OpenRouter routes endpoints at runtime + honors ZDR
|
|
6
|
+
* via a request flag, and no fleet needs per-endpoint region governance yet). NB until then `gov.region`/`dataRetention`
|
|
7
|
+
* are per-MODEL and stay UNKNOWN (fail-closed) — do NOT populate a "representative" region (it would silently degrade
|
|
8
|
+
* fail-closed to fail-OPEN at the endpoint layer — a forged in-region attestation; see C030).
|
|
9
|
+
* Decidable OpenRouter facts are NUMBERS/BOOLS; noisy third-party benchmarks are COARSE TIERS (frontier/strong/mid/basic/
|
|
4
10
|
* unknown) — never a 2-decimal score (that launders noisy/contaminated public data as precision). Every cell carries
|
|
5
11
|
* {source, asOf}; an unsourced cell is MISSING, never a confident value, and a missing tier is NEVER imputed to
|
|
6
12
|
* worst (that would kill new models). The catalog stores NO cross-axis composite — blending is the selector's job
|