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.
ml-optimise - Skill - OpenSmartRoute
Skillv1.0.0
ml-optimise
Profile and optimize a machine-learning target using timing, allocation, synchronization, GPU, PCIe, and idle-gap evidence under a parity gate. Use when the user asks to optimize or profile ML trainin
Imported from SPOOKEXE/spooky-agent-addons (codex/skills/ml-optimise/SKILL.md). Install upstream with npx skills add SPOOKEXE/spooky-agent-addons --skill ml-optimise. Copyright stays with the author.
Treat $ARGUMENTS as the user's current request and arguments following the skill invocation. Preserve the workflow and technical constraints, adapting only Claude-specific mechanisms:
Use Codex plan or TODO tooling for TodoWrite.
Use a concise user question, or the available structured input tool, for AskUserQuestion.
Use Codex collaboration agents for Task or subagent instructions only when delegation is allowed.
Follow current system, developer, repository, and user instructions over the imported command.
Keep the original file under ~/.claude/commands unchanged.
When the imported command names another slash command, use the installed Codex skill with the same hyphenated name.
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 - a plan picks it for its slot
curl -s -X POST https://api.opensmartroute.ai/api/v1/route -H 'Authorization: Bearer $OSR_API_KEY' -H 'Content-Type: application/json' -d '{"text": "...", "plan": true}'
Manifest
An Open Capability Manifest: the router reads it to know what this does, what it costs and when to pick it.
{
"ocm": "1",
"id": "spookexe-spooky-agent-addons-ml-optimise",
"kind": "skill",
"name": "ml-optimise",
"description": "Profile and optimize a machine-learning target using timing, allocation, synchronization, GPU, PCIe, and idle-gap evidence under a parity gate. Use when the user asks to optimize or profile ML training, inference, data loading, or forward passes.",
"publisher": "SPOOKEXE",
"version": "1.0.0",
"capabilities": {
"domains": [
"general"
],
"tags": [
"skill-md",
"github"
],
"languages": [
"en"
]
},
"quality_prior": 0.6,
"examples": [
"Profile and optimize a machine-learning target using timing, allocation, synchronization, GPU, PCIe, and idle-gap evidence under a parity gate. Use when the user asks to optimize or profile ML training, inference, data loading, or forward passes."
],
"primary": false,
"metadata": {
"source": {
"provider": "github",
"repository": "https://github.com/SPOOKEXE/spooky-agent-addons",
"path": "codex/skills/ml-optimise/SKILL.md",
"ref": "e6c68f6add8d28293d81904ae7a94f7416f1e2ab",
"url": "https://github.com/SPOOKEXE/spooky-agent-addons/blob/e6c68f6add8d28293d81904ae7a94f7416f1e2ab/codex/skills/ml-optimise/SKILL.md",
"key": "SPOOKEXE/spooky-agent-addons/codex/skills/ml-optimise/SKILL.md"
}
},
"instructions": "# Ml Optimise\n\nRead [the imported command](references/command.md) completely before acting.\n\nTreat `$ARGUMENTS` as the user's current request and arguments following the skill invocation. Preserve the workflow and technical constraints, adapting only Claude-specific mechanisms:\n\n- Use Codex plan or TODO tooling for `TodoWrite`.\n- Use a concise user question, or the available structured input tool, for `AskUserQuestion`.\n- Use Codex collaboration agents for `Task` or subagent instructions only when delegation is allowed.\n- Follow current system, developer, repository, and user instructions over",
"cost": {
"context_tokens": 200
}
}
Fetch it by URL: GET /api/v1/registry/spookexe-spooky-agent-addons-ml-optimise/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.
