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.

localization - Skill - OpenSmartRoute

Name: localization
Author: robotics-team-1213

Imported from robotics-team-1213/rosy-platform (operations/fleet/fleet/localization/AGENTS.md). Install upstream with npx skills add robotics-team-1213/rosy-platform --skill localization. Copyright stays with the author.

localization

Parent context: ../AGENTS.md Updated: 2026-10-14

Purpose

D-395 Fleet localization arbiter. Scores each robot's pose candidates with bounded cues and decides only when one leads the next by a margin held for 2 s. Pure: no transport, no asyncio, no robot calls. The service loop that feeds it and sends decisions belongs to server/ (Phase 2).

Key Files

File	Description
`__init__.py`	Package marker
`cues.py`	Bounded cues per hypothesis: peers, slot (10 cm / 20 deg, either way along the axis), last good pose (never after a pickup), overhead sighting (≤ 300 ms), reference square
`arbiter.py`	`Weights`, `Context`, `score()`, `Arbiter.observe()` → `LocalizationDecision`
(not in this folder)	MapPose lives in `rosy.world.state.projections.map_pose`. Import it directly (D-752). This folder does not re-export it. Fleet service, transaction, and admission stay in `server/`
`lane_compliance.py`	D-511 2: signed lateral offset (left +), body margin `width_m/2 − (

For AI Agents

Working In This Directory

Keep it pure: test_boundaries.py forbids httpx, websockets, rclpy, asyncio, fastapi, fleet.swarm and fleet.server here.
A missing input is a 0 cue, never a guess; the arbiter must work with no camera and no peers.
last_good and overhead weigh less than the margin on purpose: neither may decide alone (D-395 §7).

Testing Requirements

python -m pytest operations/fleet/test/test_localization_cues.py operations/fleet/test/test_localization_arbiter.py operations/fleet/test/test_boundaries.py -q

Dependencies

Internal

core_common.protocol.localization (CandidateReport, LocalizationDecision)

External

None.

Install localization 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": "robotics-team-1213-rosy-platform-localization", "kind": "skill", "name": "localization", "description": "D-395 Fleet localization arbiter. Scores each robot's pose candidates with bounded cues and decides only when one leads the next by a margin held for 2 s. Pure: no transport, no asyncio, no robot calls. The service loop that feeds it and sends decisions belongs to `server/` (Phase 2).", "publisher": "robotics-team-1213", "version": "1.0.0", "capabilities": { "domains": [ "coding", "science" ], "tags": [ "skill-md", "github-agents" ], "languages": [ "en" ] }, "quality_prior": 0.6, "examples": [ "D-395 Fleet localization arbiter. Scores each robot's pose candidates with bounded cues and decides only when one leads the next by a margin held for 2 s. Pure: no transport, no asyncio, no robot calls. The service loop that feeds it and sends decisions belongs to `server/` (Phase 2)." ], "primary": false, "metadata": { "source": { "provider": "github-agents", "repository": "https://github.com/robotics-team-1213/rosy-platform", "path": "operations/fleet/fleet/localization/AGENTS.md", "ref": "c47d7c3e78c9751a9428450579bb186af0520725", "url": "https://github.com/robotics-team-1213/rosy-platform/blob/c47d7c3e78c9751a9428450579bb186af0520725/operations/fleet/fleet/localization/AGENTS.md", "key": "robotics-team-1213/rosy-platform/operations/fleet/fleet/localization/AGENTS.md" } }, "instructions": "\n\n\n# localization\n\n**Parent context:** `../AGENTS.md`\n**Updated:** 2026-10-14\n\n## Purpose\n\nD-395 Fleet localization arbiter. Scores each robot's pose candidates with bounded cues and decides only when one leads the next by a margin held for 2 s. Pure: no transport, no asyncio, no robot calls. The service loop that feeds it and sends decisions belongs to `server/` (Phase 2).\n\n## Key Files\n\n| File | Description |\n|------|-------------|\n| `__init__.py` | Package marker |\n| `cues.py` | Bounded cues per hypothesis: pee", "cost": { "context_tokens": 514 } }

localization

About

localization

Purpose

Key Files

For AI Agents

Working In This Directory

Testing Requirements

Dependencies

Internal

External

Install localization 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

Use it

Manifest

Reviews