agentsclimarketplace

Model recommender

Skill navveenb/lean-agentic-ai/lean-agentic-ai-claude-implementation/.claude/skills/model-recommender

Lean Agentic AI

Install
npx -y skills add navveenb/lean-agentic-ai --skill model-recommender

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

2 things to look at

  • no licenseNo license file was found in the repository. Code published without one is not open source by default, so using it at work is a question for whoever answers licensing questions where you are.
  • 9 stars9 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

What its author says it does

Copied from the file, not written here

Recommend the right model tier for each LLM task type

SKILL.md

1.5 KB, 359 tokens by cl100k_base, as published. Nobody here has run it

Model Recommender Skill

Task-to-Tier Mapping

Tier 1: Small / Fast (Haiku, GPT-4o-mini, Gemini Flash)

  • Classification (spam/not-spam, category assignment)
  • Entity extraction (names, dates, emails)
  • Sentiment analysis
  • Routing decisions
  • Simple formatting (JSON restructuring, template filling)
  • Yes/no judgments, slot filling

Tier 2: Mid-Range (Sonnet, GPT-4o, Gemini Pro)

  • Summarization (single document)
  • Translation
  • Simple code generation (known patterns)
  • Structured data extraction from complex documents
  • Q&A with provided context
  • Email/message drafting

Tier 3: Frontier (Opus, GPT-4/o1, Gemini Ultra)

  • Complex multi-step reasoning
  • Novel problem solving
  • Nuanced creative writing
  • Multi-document synthesis
  • Architectural planning
  • Novel/complex code generation

Decision Rules

  1. When in doubt, start small — default to Tier 1.
  2. Classification is ALWAYS Tier 1.
  3. Extraction is almost always Tier 1.
  4. Summarization is Tier 2 (unless multi-document → Tier 3).
  5. Code generation: boilerplate = Tier 1, standard = Tier 2, novel = Tier 3.
  6. If a task could be done with regex/rules → no LLM at all.

Red Flags

  • Frontier model for classification → should be Tier 1
  • Frontier model for formatting/extraction → should be Tier 1
  • Frontier model for simple summarization → should be Tier 2
  • Any LLM call for a deterministic task → should be code, not LLM

What ships with it

Read from the repository

Just SKILL.md. No reference files, no scripts.

Keep looking

Skills are one crate of 328,083. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.