Imported from AnkitClassicVision/ankit_shared_skills (caliber/SKILL.md). Install upstream with npx skills add AnkitClassicVision/ankit_shared_skills --skill caliber. Copyright stays with the author (MIT).

CALIBER

Name: caliber
Author: AnkitClassicVision

CALIBER is a model-routing skill. It turns a request into a task-surface demand vector, compares that vector against a model capability matrix, and returns the cheapest constructor that can clear the task with the right defender.

A constructor is:

model + deployment path + context + retrieval + tools + verifier + human gate

Core rule

Do not ask “what is the best model?” Ask:

What is the cheapest allowed constructor whose proven capabilities clear this task's required surfaces and whose defender can verify the result?

Demand levels

Level	Name	Meaning
-1	deterministic	No model; use SQL/rules/regex/script/API directly.
0	tiny	Simple classify/extract/rewrite/routing.
1	small	Simple reasoning, short summaries, low-stakes drafts.
2	workhorse	Normal professional work, routine code/specs, common RAG.
3	strong specialist	Harder domain work, repo work, complex synthesis, multimodal reasoning.
4	frontier	Expensive-to-fail planning, novel architecture, high-reliability synthesis.
5	orchestrated frontier	Decomposition + tools + defenders + human gates.
6	reserved future	Future super-frontier/research-grade tier.

Workflow

Gate the deployment pool: regulated/private data, local-only requirements, secrets, production writes, or external sends may restrict allowed models.
Break the task into surfaces such as reasoning, coding, long context, tool use, media, structured output, local/private, or compliance.
Assign required level per surface.
Filter candidate models/deployments from references/model-capability-matrix.yaml.
Attach defenders from references/defender-registry.md.
Pick the cheapest viable constructor.
If the defender fails, escalate one tier or switch specialist lane.

Output shape

Task surfaces:
- <surface>: required level <n> — <why>

Gate pool:
- allowed deployments:
- excluded deployments:

Candidate comparison:
- cheapest viable:
- strongest reliable:
- local/private option:
- specialist option:

Defenders:
- required verifier(s):
- human gate:

Recommendation:
- primary constructor:
- fallback constructor:
- escalation rule:
- evidence confidence:

References

references/surface-taxonomy.md
references/model-capability-matrix.yaml
references/model-capability-schema.json
references/benchmark-source-ledger.md
references/defender-registry.md
references/routing-policy-v0.3.md
references/model-registry.md
references/demand-card.md

{ "ocm": "1", "id": "ankitclassicvision-ankit-shared-skills-caliber", "kind": "skill", "name": "caliber", "description": "Route prompts by task-demand surface before spending tokens. Use when deciding which model, tool lane, verifier, or human gate a task needs; when a task involves local/private models, small/cheap models, coding, agents, long context, media, compliance, or high-stakes verification.", "publisher": "AnkitClassicVision", "version": "1.0.0", "capabilities": { "domains": [ "legal" ], "tags": [ "skill-md", "github" ], "languages": [ "en" ] }, "quality_prior": 0.6, "examples": [ "Route prompts by task-demand surface before spending tokens. Use when deciding which model, tool lane, verifier, or human gate a task needs; when a task involves local/private models, small/cheap models, coding, agents, long context, media, compliance, or high-stakes verification." ], "primary": false, "metadata": { "source": { "provider": "github", "repository": "https://github.com/AnkitClassicVision/ankit_shared_skills", "path": "caliber/SKILL.md", "ref": "f986c6d5de8a3591b9d9c1fad21ddc959528a159", "url": "https://github.com/AnkitClassicVision/ankit_shared_skills/blob/f986c6d5de8a3591b9d9c1fad21ddc959528a159/caliber/SKILL.md", "key": "AnkitClassicVision/ankit_shared_skills/caliber/SKILL.md" }, "license": "MIT" }, "instructions": "# CALIBER\n\nCALIBER is a model-routing skill. It turns a request into a **task-surface demand\nvector**, compares that vector against a **model capability matrix**, and returns\nthe cheapest constructor that can clear the task with the right defender.\n\nA constructor is:\n\n```text\nmodel + deployment path + context + retrieval + tools + verifier + human gate\n```\n\n## Core rule\n\nDo not ask “what is the best model?” Ask:\n\n> What is the cheapest allowed constructor whose proven capabilities clear this\n> task's required surfaces and whose defender can verify the result?\n\n## Demand levels\n\n| Level | Name ", "cost": { "context_tokens": 646 } }

caliber

About

CALIBER

Core rule

Demand levels

Workflow

Output shape

References

Use it

Manifest

Reviews