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.

AI_Avatar - Skill - OpenSmartRoute

Imported from thangngny/AI_Avatar (AGENTS.md). Install upstream with npx skills add thangngny/AI_Avatar. Copyright stays with the author.

AGENTS.md

MANDATORY STARTUP

Before modifying or running this repository:

Confirm the Git branch is main (git branch --show-current).
Read SYSTEM_CONTROL_BOARD.md (the shared system board).
Read config/render_truth.json (the only runtime quality/lip-sync truth).
Treat render_truth.json as the single source of truth for locked render parameters. Do not reintroduce quality defaults elsewhere.
avatar_id, script, master audio, and output path are variable job inputs only — never part of the locked truth.
Never confuse a RAW avatar source with a PREPARED avatar.
- RAW source = the original camera .mp4/.mov/photo before MuseTalk preprocessing → data/input/avatar_sources/ (and lineage index data/avatar_sources.json).
- PREPARED avatar = data/avatars/<avatar_id>/ (full_imgs/, coords.pkl, latents.pt, mask/, mask_coords.pkl, avator_info.json) — the generated asset the renderer actually consumes.
- A data/avatars/.uploads/*.mp4 is a RAW source, not a prepared avatar.
Do not use experimental pipelines (experiments/, archive/, the LatentSync copy) unless explicitly requested. Production runtime is MuseTalk v1.5 only.
If not on main, STOP and return to main before normal production work.
Preflight (optional, no models loaded): python system_status.py.

Do not duplicate the control board here. SYSTEM_CONTROL_BOARD.md is authoritative; this file only points to it.

Install AI_Avatar 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.

AI_Avatar

About

AGENTS.md

MANDATORY STARTUP

Use it

Install AI_Avatar 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