How long does the integration take?

Change one base URL in an OpenAI-compatible client, or install the Python or TypeScript SDK; the quickstart routes a first request in a few minutes.

Will this lock us into another vendor?

No. The router is Apache-2.0 open source, your provider keys stay yours, and the API is OpenAI-compatible, so leaving is a change of base URL.

What happens when a provider fails?

The router falls back to the next model your policy allows and records the fallback in the trace, so an outage costs a retry, not an incident.

bayes - Skill - OpenSmartRoute

Name: bayes
Author: PamelaSprin47685ghall

Imported from PamelaSprin47685ghall/wanxiangli (src/bayes/AGENTS.md). Install upstream with npx skills add PamelaSprin47685ghall/wanxiangli --skill bayes. Copyright stays with the author.

bayes

层：纯数值。Sphinx v2 有限离散假设集上精确贝叶斯后验的唯一 owner；不得 import 宿主 SDK，不调用模型、不读写存储、不把后验说成实际正确率。npm 入口 wanxiangli/bayes。

AXIOM

[使用者] 只有合格因子进入后验；因子必须声明依赖组、覆盖全部假设、似然有限且在 [0,1]（sphinx-v2-009/025）。
[使用者] 同一观测 id 重复投递只算一次；同组不同 id 不得被静默吞掉，未声明联合因子时只保留首个并逐条记录被丢弃者（sphinx-v2-021）。
[使用者] 对数空间乘积与 log-sum-exp 归一，极小尺度不下溢，零先验保持为零；后验只在声明模型内成立（sphinx-v2-026）。

LEMMA

[AXIOM 1] 假设少于 2、空键、重复键（报出真正重复的键）、非法先验、零先验质量、无合格因子、空依赖组、未知键、缺键、越界似然各有类型化故障。
[AXIOM 2] 同 id 重复不改变后验；同组不同 id 时 dropped 列出其余 id。
[AXIOM 3] 后验与闭式一致；似然 1e-300 量级仍归一；零先验后验为 0；全零似然报 NonPositivePartition。
[AXIOM 1] priorOnly 返回归一先验、usedObservations 为空，且与 infer 共用假设集契约。

SPEC

[LEMMA 1–4] 证明：tests/001.test.mjs。

# after Install: the listing is in your workspace's routing pool - a plan picks it for its slot curl -s -X POST https://api.opensmartroute.ai/api/v1/route -H 'Authorization: Bearer $OSR_API_KEY' -H 'Content-Type: application/json' -d '{"text": "...", "plan": true}'

Install bayes and it becomes one more candidate the router can pick - when it fits.

A listing is a routing target with a manifest: what it does, which domains it covers, what it costs and who publishes it. Once installed it sits beside your own models and tools, is scored like any other candidate for each request, and shows up in the trace when it wins. Ratings come from workspaces that installed it, one per account.

{ "ocm": "1", "id": "pamelasprin47685ghall-wanxiangli-bayes", "kind": "skill", "name": "bayes", "description": "层：纯数值。Sphinx v2 有限离散假设集上精确贝叶斯后验的唯一 owner；不得 import 宿主 SDK，不调用模型、不读写存储、不把后验说成实际正确率。npm 入口 `wanxiangli/bayes`。", "publisher": "PamelaSprin47685ghall", "version": "1.0.0", "capabilities": { "domains": [ "coding", "math" ], "tags": [ "skill-md", "github-agents" ], "languages": [ "en" ] }, "quality_prior": 0.6, "examples": [ "层：纯数值。Sphinx v2 有限离散假设集上精确贝叶斯后验的唯一 owner；不得 import 宿主 SDK，不调用模型、不读写存储、不把后验说成实际正确率。npm 入口 `wanxiangli/bayes`。" ], "primary": false, "metadata": { "source": { "provider": "github-agents", "repository": "https://github.com/PamelaSprin47685ghall/wanxiangli", "path": "src/bayes/AGENTS.md", "ref": "663ba59b8282db5e3eeccda0c1100ea8cdca0dd0", "url": "https://github.com/PamelaSprin47685ghall/wanxiangli/blob/663ba59b8282db5e3eeccda0c1100ea8cdca0dd0/src/bayes/AGENTS.md", "key": "PamelaSprin47685ghall/wanxiangli/src/bayes/AGENTS.md" } }, "instructions": "# bayes\n\n层：纯数值。Sphinx v2 有限离散假设集上精确贝叶斯后验的唯一 owner；不得 import 宿主 SDK，不调用模型、不读写存储、不把后验说成实际正确率。npm 入口 `wanxiangli/bayes`。\n\n# AXIOM\n\n1. [使用者] 只有合格因子进入后验；因子必须声明依赖组、覆盖全部假设、似然有限且在 [0,1]（sphinx-v2-009/025）。\n2. [使用者] 同一观测 id 重复投递只算一次；同组不同 id 不得被静默吞掉，未声明联合因子时只保留首个并逐条记录被丢弃者（sphinx-v2-021）。\n3. [使用者] 对数空间乘积与 log-sum-exp 归一，极小尺度不下溢，零先验保持为零；后验只在声明模型内成立（sphinx-v2-026）。\n\n# LEMMA\n\n1. [AXIOM 1] 假设少于 2、空键、重复键（报出真正重复的键）、非法先验、零先验质量、无合格因子、空依赖组、未知键、缺键、越界似然各有类型化故障。\n2. [AXIOM 2] 同 id 重复不改变后验；同组不同 id 时 `dropped` 列出其余 id。\n3. [AXIOM 3] 后验与闭式一致；似然 1e-300 量级仍归一；零先验后验为 0；全零似然报 NonPositivePartition。\n4. [AXIOM 1] `priorOnly` 返回", "cost": { "context_tokens": 173 } }

bayes

About

bayes

AXIOM

LEMMA

SPEC

Use it

Install bayes and it becomes one more candidate the router can pick - when it fits.

What it costs to adopt - in engineering time.

One base URL

No vendor lock-in

Failover built in

Manifest

Reviews