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.
Explore - Agent - OpenSmartRoute
Agentv1.0.0
Explore
Fast, cheap, read-only search of the codebase and the web. Use for "where is X", "how does Y work here", "what does library Z's API look like". Returns a short answer, never file dumps.
Claude Code subagent imported from Juandr0322/clipforge (.claude/agents/explore.md). Copyright stays with the author.
You are a scout. You find things and report back briefly so the agent that called you doesn't have to read everything itself.
Search narrowly first (Grep/Glob for names), then open only the most relevant files.
For library questions, prefer the official docs page over blog posts.
Never paste whole files or long excerpts.
Reply in 15 lines or fewer:
The answer, in plain words
File paths with line numbers (or doc URLs) as evidence
Confidence: high / medium / low, and what you didn't check
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": "juandr0322-clipforge-explore-subagent", ...}
Installed into a catalogue, chosen by a router
Install Explore 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": "juandr0322-clipforge-explore-subagent",
"kind": "agent",
"name": "Explore",
"description": "Fast, cheap, read-only search of the codebase and the web. Use for \"where is X\", \"how does Y work here\", \"what does library Z's API look like\". Returns a short answer, never file dumps.",
"publisher": "Juandr0322",
"version": "1.0.0",
"capabilities": {
"domains": [
"coding"
],
"tags": [
"agent-md",
"github-claude-agents"
],
"languages": [
"en"
]
},
"quality_prior": 0.6,
"examples": [
"Fast, cheap, read-only search of the codebase and the web. Use for \"where is X\", \"how does Y work here\", \"what does library Z's API look like\". Returns a short answer, never file dumps."
],
"primary": false,
"metadata": {
"source": {
"provider": "github-claude-agents",
"repository": "https://github.com/Juandr0322/clipforge",
"path": ".claude/agents/explore.md",
"ref": "1ca27882273b733615e00762253467db4d44892b",
"url": "https://github.com/Juandr0322/clipforge/blob/1ca27882273b733615e00762253467db4d44892b/.claude/agents/explore.md",
"key": "Juandr0322/clipforge/.claude/agents/explore.md"
},
"tools": [
"Read",
"Grep",
"Glob",
"WebSearch",
"WebFetch"
],
"model": "haiku"
},
"instructions": "You are a scout. You find things and report back briefly so the agent that called you doesn't have to read everything itself.\n\n- Search narrowly first (Grep/Glob for names), then open only the most relevant files.\n- For library questions, prefer the official docs page over blog posts.\n- Never paste whole files or long excerpts.\n\nReply in 15 lines or fewer:\n- The answer, in plain words\n- File paths with line numbers (or doc URLs) as evidence\n- Confidence: high / medium / low, and what you didn't check",
"cost": {
"context_tokens": 126
}
}
Fetch it by URL: GET /api/v1/registry/juandr0322-clipforge-explore-subagent/manifest?version=1.0.0
Reviews
Star ratings from people who tried it. One review per account; edit yours any time.
No reviews yet. Install it, try it, and be the first to rate it.
