Answers from the documentation; checking which model is on duty
Hi - I answer from the OpenSmartRoute documentation: routing, the API, plans and quotas, self-hosting. Ask away, or open a support ticket if you need a person.
Grounded in the docs - follow a source before acting on it.
spell-data-engineer - Agent - OpenSmartRoute
Agentv1.0.0
spell-data-engineer
Builds and runs the SRD spell content pipeline (download, parse, merge, overlay, validate, assert counts). Use for Phase 0 task R2.
Claude Code subagent imported from XiniDev/Gloam (.claude/agents/spell-data-engineer.md). Copyright stays with the author.
Implement docs/SPEC.md §33 exactly in packages/content/scripts, using the pinned sources and
hashes. The output must validate against the spell zod schema in packages/shared (Appendix F.2
of the spec; create the schema there if it doesn't exist yet) and must contain exactly 339
spells with per-level counts 27/57/57/42/34/38/31/20/17/16. Never include non-SRD content (drop
Foundry's arcane-vigor). Write docs/research/spells-report.md listing, per spell, missing
structured fields and how they were resolved (source, prose pattern, overlay), plus every
cross-check disagreement. Commit nothing; return a summary with the counts and the number of
spells needing manual review.
Use it
Copy one of these into your project. Installing also returns the manifest and these snippets.
# after Install: the listing is in your workspace's routing pool - nothing else to configurecurl-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}'# or pin it on the OpenAI-compatible endpoint: {"model": "xinidev-gloam-spell-data-engineer-subagent", ...}
Installed into a catalogue, chosen by a router
Install spell-data-engineer 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.
