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.

next - Prompt - OpenSmartRoute

Name: next
Author: albegali

90 days of the enterprise, pro and team plans free, no card needed. See the plans

OpenSmartRoute

Promptv1.0.0

Compute the next workflow step and tell which command, mode and model to use (minimal cost)

by albegali(0) 0 installs

Free

Free account. Installing gives you the manifest plus copy-paste snippets.

See reviews

About

Prompt file imported from albegali/agentic-skeleton (.codex/prompts/next.md). Fill in {{arguments}} before use. Copyright stays with the author.

Recommended model: gpt-5-mini (tier fast, see models.toml). Switch to it before running this command.

Do not explore the repository: the state is computed by the script.

Run python3 tools/next_step.py {{arguments}}.
Read only workflow/steps/<recommended step>.md (and the optional parallel step's file, if any). Nothing else.
Answer in at most 10 lines:

State: <phase / current sprint, 1 line>
Next: <step> — <IDs>
Who: <mode | YOU (human gate)>  ·  Tier: <tier> → <model printed by the script>
Why: <1 line>
Run: <fresh session> → <exact command with IDs>     (or: "human action: …")
In parallel (optional): <command>
Waiting on you: <human gates, if any>

Do not execute the step in this session. If the weekly reminder is printed, add one line with /weekly.

Use it

Copy one of these into your project. Installing also returns the manifest and these snippets.

<!-- Generated by tools/render_agents.py from workflow/workflow.toml and models.toml — do not edit; run `make agents`. -->
> Recommended model: **gpt-5-mini** (tier `fast`, see models.toml). Switch to it before running this command.

Do not explore the repository: the state is computed by the script.

1. Run `python3 tools/next_step.py {{arguments}}`.
2. Read only `workflow/steps/<recommended step>.md` (and the optional parallel step's file, if any). Nothing else.
3. Answer in at most 10 lines:

```
State: <phase / current sprint, 1 line>
Next: <step> — <IDs>
Who: <mode | YOU (human gate)>  ·  Tier: <tier> → <model printed by the script>
Why: <1 line>
Run: <fresh session> → <exact command with IDs>     (or: "human action: …")
In parallel (optional): <command>
Waiting on you: <human gates, if any>
```

4. Do not execute the step in this session. If the weekly reminder is printed, add one line with `/weekly`.

Installed into a catalogue, chosen by a router

Install next 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.

Photo: RDNE Stock project

For the business

What it costs to adopt - in engineering time.

Same client, same code
One base URL
The OpenAI-compatible endpoint answers your existing SDK calls; model=auto lets the router choose.
The SDKs
One URL, every vendor
No vendor lock-in
Point your OpenAI-compatible client at one address; add, swap or retire providers in the catalogue without touching application code.
Every vendor side by side
Next eligible model answers
Failover built in
When a provider is down or over quota the request moves on; the trace says it did.

Manifest

The prompt text and its fill-in variables. Copy it or fetch it by URL from your own code.

albegali-agentic-skeleton-next-codex-prompt.prompt.jsonjson

{
  "prompt": "<!-- Generated by tools/render_agents.py from workflow/workflow.toml and models.toml — do not edit; run `make agents`. -->\n> Recommended model: **gpt-5-mini** (tier `fast`, see models.toml). Switch to it before running this command.\n\nDo not explore the repository: the state is computed by the script.\n\n1. Run `python3 tools/next_step.py {{arguments}}`.\n2. Read only `workflow/steps/<recommended step>.md` (and the optional parallel step's file, if any). Nothing else.\n3. Answer in at most 10 lines:\n\n```\nState: <phase / current sprint, 1 line>\nNext: <step> — <IDs>\nWho: <mode | YOU (human gate)>  ·  Tier: <tier> → <model printed by the script>\nWhy: <1 line>\nRun: <fresh session> → <exact command with IDs>     (or: \"human action: …\")\nIn parallel (optional): <command>\nWaiting on you: <human gates, if any>\n```\n\n4. Do not execute the step in this session. If the weekly reminder is printed, add one line with `/weekly`.",
  "variables": [
    "arguments"
  ],
  "notes": "Compute the next workflow step and tell which command, mode and model to use (minimal cost) Arguments: [ID]  (optional: state of one story).",
  "metadata": {
    "source": {
      "provider": "github-codex-prompts",
      "repository": "https://github.com/albegali/agentic-skeleton",
      "path": ".codex/prompts/next.md",
      "ref": "ad43875e26123ba40eed52c994c37d3c9458e658",
      "url": "https://github.com/albegali/agentic-skeleton/blob/ad43875e26123ba40eed52c994c37d3c9458e658/.codex/prompts/next.md",
      "key": "albegali/agentic-skeleton/.codex/prompts/next.md"
    }
  }
}

Fetch it by URL: GET /api/v1/registry/albegali-agentic-skeleton-next-codex-prompt/manifest?version=1.0.0

About

Use it

Install next 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