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.

/code-review - Prompt - OpenSmartRoute

Prompt file imported from fairagro/m4.2_sql_to_arc (.cursor/commands/code-review.md). Copyright stays with the author.

code-review

Produce a first-party critical review of a local base...HEAD diff (default base main) or a GitHub pull request.

Output: write a report under /tmp via m42-ai (body starts with , numbered findings with Path bullets — no Findings index table); for a PR, publish a formal COMMENT Pull Request Review when auth works. Do not triage Copilot/Bugbot threads (use /review-fixer). Do not commit, push, or auto-approve.

Input: optional base ref, and/or PR number/URL.

Steps

Read and follow .agents/skills/code-review/SKILL.md.
Use docs/ai_review_policy.md for severity language; do not duplicate Ruff/mypy/Bandit/etc. findings.
Prefer uv run --project scripts/ai m42-ai code-review-context|code-review-report-write|code-review-publish.
If publishing to a PR and GH_TOKEN is missing with no TTY, ask the user to source ./scripts/set-dev-tokens.sh and wait — keep the /tmp report if auth fails.

# code-review Produce a first-party critical review of a local `base...HEAD` diff (default base `main`) or a GitHub pull request. **Output:** write a report under `/tmp` via `m42-ai` (body starts with ``, numbered findings with **Path** bullets — no Findings index table); for a PR, publish a formal COMMENT Pull Request Review when auth works. Do not triage Copilot/Bugbot threads (use `/review-fixer`). Do not commit, push, or auto-approve. **Input:** optional base ref, and/or PR number/URL. **Steps** 1. Read and follow `.agents/skills/code-review/SKILL.md`. 2. Use `docs/ai_review_policy.md` for severity language; do not duplicate Ruff/mypy/Bandit/etc. findings. 3. Prefer `uv run --project scripts/ai m42-ai code-review-context|code-review-report-write|code-review-publish`. 4. If publishing to a PR and `GH_TOKEN` is missing with no TTY, ask the user to `source ./scripts/set-dev-tokens.sh` and wait — keep the `/tmp` report if auth fails.

Install /code-review 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.

/code-review

About

code-review

Use it

Install /code-review 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