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.
See `content/06-decision-tree.xml`. The tree picks retriever set based on corpus type (free-form prose, structured docs, code) and metric set based on task (retrieval vs generation). Walk it before dr
Imported from faionfaion/faion-network (skills/faion/knowledge/ai-core/rag-bench-harness-template/AGENTS.md). Install upstream with npx skills add faionfaion/faion-network --skill rag-bench-harness-template. Copyright stays with the author.
RAG Bench Harness Template
Summary
One-sentence: Spec for a reusable RAG bench harness — corpus snapshot, query set, retriever runners, eval metrics, leaderboard schema — so chunking-strategy comparisons stop being hand-wave.
One-paragraph: Existing RAG eval methodologies are atomic (rag-eval-metrics, rag-eval-ab-testing). A wire-it-together harness that runs a chunking experiment end-to-end is missing. This methodology produces a rag-bench-spec.json artefact pinning corpus snapshot, query set, retriever runners, eval metrics (Recall@k, MRR, faithfulness), and a leaderboard schema. Output is a versioned harness consumed by chunking-strategy A/Bs.
Ефективно для:
Chunking strategy bench (recurring, weekly or quarterly).
Compare retrievers (BM25 vs dense vs hybrid) on the same corpus.
Reproducible eval — corpus snapshot pinned by sha.
Leaderboard schema для cross-strategy comparison.
Bridge до downstream [[prompt-ab-power-calculator]] для chunking A/B.
Narrative review draft. Generated from templates/rag-bench-spec.md.j2 by tpl-jinja --migrate; do not hand-edit.
Scripts
File
Purpose
When to call
scripts/validate-rag-bench-harness-template.py
Validate rag-bench-spec
Pre-commit + before bench run
Related
[[prompt-ab-power-calculator]]
[[rag-corpus-discovery-interview]]
[[production-trace-mining-for-training-data]]
Decision tree
See content/06-decision-tree.xml. The tree picks retriever set based on corpus type (free-form prose, structured docs, code) and metric set based on task (retrieval vs generation). Walk it before drafting the spec.
Use it
Copy one of these into your project. Installing also returns the manifest and these snippets.
yaml
targets:
- https://api.opensmartroute.ai/api/v1/registry/faionfaion-faion-network-rag-bench-harness-template/manifest # or paste the manifest below
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": "faionfaion-faion-network-rag-bench-harness-template",
"kind": "skill",
"name": "rag-bench-harness-template",
"description": "See `content/06-decision-tree.xml`. The tree picks retriever set based on corpus type (free-form prose, structured docs, code) and metric set based on task (retrieval vs generation). Walk it before drafting the spec.",
"publisher": "faionfaion",
"version": "1.0.0",
"capabilities": {
"domains": [
"coding"
],
"tags": [
"skill-md",
"github-agents"
],
"languages": [
"en"
]
},
"quality_prior": 0.6,
"examples": [
"See `content/06-decision-tree.xml`. The tree picks retriever set based on corpus type (free-form prose, structured docs, code) and metric set based on task (retrieval vs generation). Walk it before drafting the spec."
],
"primary": false,
"metadata": {
"source": {
"provider": "github-agents",
"repository": "https://github.com/faionfaion/faion-network",
"path": "skills/faion/knowledge/ai-core/rag-bench-harness-template/AGENTS.md",
"ref": "8e31684537e64d45c75264b1e0a2384f9b024a34",
"url": "https://github.com/faionfaion/faion-network/blob/8e31684537e64d45c75264b1e0a2384f9b024a34/skills/faion/knowledge/ai-core/rag-bench-harness-template/AGENTS.md",
"key": "faionfaion/faion-network/skills/faion/knowledge/ai-core/rag-bench-harness-template/AGENTS.md"
}
},
"instructions": "# RAG Bench Harness Template\n\n## Summary\n\n**One-sentence:** Spec for a reusable RAG bench harness — corpus snapshot, query set, retriever runners, eval metrics, leaderboard schema — so chunking-strategy comparisons stop being hand-wave.\n\n**One-paragraph:** Existing RAG eval methodologies are atomic (rag-eval-metrics, rag-eval-ab-testing). A wire-it-together harness that runs a chunking experiment end-to-end is missing. This methodology produces a `rag-bench-spec.json` artefact pinning corpus snapshot, query set, retriever runners, eval metrics (Recall@k, MRR, faithfulness), and a leaderboard s",
"cost": {
"context_tokens": 963
}
}
Fetch it by URL: GET /api/v1/registry/faionfaion-faion-network-rag-bench-harness-template/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.