Skip to main content
skill CX Codex

llm-tuning-patterns — Claude Code skill: settings and install

The llm-tuning-patterns skill recommends LLM settings for theorem proving, code generation, and creative work. Use it to choose token limits and temperature for a specific task.

MIT ★ 3,951 updated 2026-01-26

Install into Claude Code · Skills CLI (Vercel) All methods and platforms ↓
npx skills add parcadei/Continuous-Claude-v3 --skill llm-tuning-patterns -a claude-code
Open on GitHub

What llm-tuning-patterns does

The skill gives separate parameter suggestions for different workloads. For theorem proving, it recommends max_tokens 4096, temperature 0.6, and top_p 0.95. For code generation, it suggests 2048 tokens and a temperature between 0.2 and 0.4; creative tasks call for 4096 tokens and a temperature from 0.8 to 1.0.

For Lean 4 proofs, it advises writing a high-level plan before selecting tactics. Hard problems can use parallel sampling: generate 8 to 32 candidates, then choose the best. A low token limit can cut off a proof, while temperature 0.2 may miss less obvious tactic paths. These are configuration recommendations; the skill does not automatically change model settings.

Install it with npx skills add parcadei/Continuous-Claude-v3 --skill llm-tuning-patterns -a claude-code. The skill file is located at ~/.claude/skills/llm-tuning-patterns/SKILL.md.

What it helps with

  • choose a temperature for code generation
  • set LLM parameters for formal proof work
  • plan a Lean 4 proof before selecting tactics
  • compare candidate attempts at a difficult proof

Example prompts for the agent

  • Recommend LLM settings for this code-generation task and explain the choice.
  • Write a proof plan for this theorem, then suggest Lean 4 tactics for each step.
  • Suggest max_tokens and temperature values for a creative exploration task.

How to install llm-tuning-patterns

1Skills CLI (Vercel)

npx skills add parcadei/Continuous-Claude-v3 --skill llm-tuning-patterns -a claude-code

2Manually: put the files in place

~/.claude/skills/llm-tuning-patterns/SKILL.md

Copy the whole skill folder; for one project use .claude/skills/llm-tuning-patterns/

Step by step, with commands for each platform: How to install skills in Claude Code: five ways

Questions about llm-tuning-patterns

How do I install llm-tuning-patterns in Claude Code?

Run npx skills add parcadei/Continuous-Claude-v3 --skill llm-tuning-patterns -a claude-code. The skill file is at ~/.claude/skills/llm-tuning-patterns/SKILL.md.

What does llm-tuning-patterns do?

It recommends LLM parameters for theorem proving, code generation, and creative tasks. For Lean 4 proofs, it also advises planning the approach before choosing tactics.

Which settings does the skill recommend for theorem proving?

It suggests max_tokens 4096, temperature 0.6, and top_p 0.95. For difficult proofs, it describes generating 8 to 32 candidates in parallel and selecting the best.

Source text

SKILL.md · MIT
---
name: llm-tuning-patterns
description: LLM Tuning Patterns
user-invocable: false
---

# LLM Tuning Patterns

Evidence-based patterns for configuring LLM parameters, based on APOLLO and Godel-Prover research.

## Pattern

Different tasks require different LLM configurations. Use these evidence-based settings.

## Theorem Proving / Formal Reasoning

Based on APOLLO parity analysis:

| Parameter | Value | Rationale |
|-----------|-------|-----------|
| max_tokens | 4096 | Proofs need space for chain-of-thought |
| temperature | 0.6 | Higher creativity for tactic exploration |
| top_p | 0.95 | Allow diverse proof paths |

### Proof Plan Prompt

Always request a proof plan before tactics:

```
Given the theorem to prove:
[theorem statement]

First, write a high-level proof plan explaining your approach.
Then, suggest Lean 4 tactics to implement each step.
```

The proof plan (chain-of-thought) significantly improves tactic quality.

### Parallel Sampling

For hard proofs, use parallel sampling:
- Generate N=8-32 candidate proof attempts
- Use best-of-N selection
- Each sample at temperature 0.6-0.8

## Code Generation

| Parameter | Value | Rationale |
|-----------|-------|-----------|
| max_tokens | 2048 | Sufficient for most functions |
| temperature | 0.2-0.4 | Prefer deterministic output |

## Creative / Exploration Tasks

| Parameter | Value | Rationale |
|-----------|-------|-----------|
| max_tokens | 4096 | Space for exploration |
| temperature | 0.8-1.0 | Maximum creativity |

## Anti-Patterns

- **Too low tokens for proofs**: 512 tokens truncates chain-of-thought
- **Too low temperature for proofs**: 0.2 misses creative tactic paths
- **No proof plan**: Jumping to tactics without planning reduces success rate

## Source Sessions

- This session: APOLLO parity - increased max_tokens 512->4096, temp 0.2->0.6
- This session: Added proof plan prompt for chain-of-thought before tactics

© parcadei · MIT · parcadei/Continuous-Claude-v3

Similar skills

All skills for AI and ML →

From the same set161

The whole set →
leann-search leann-search runs semantic code queries against a LEANN vector index when exact keyword matches fall short. CX Codex +1 ★ 3,951repo stars natural-transformations This skill guides category-theory work on natural transformations, with Lean 4 notation for checking naturality and applying related constructions. CX Codex +1 ★ 3,951repo stars categories-functors categories-functors — a skill for solving category and functor problems in category theory. Validates axioms, checks functor properties, and verifies proofs in Lean 4 via compiler. CX Codex +1 ★ 3,951repo stars limits-colimits limits-colimits is a Claude Code skill that provides decision workflows for category theory limits and colimits, including Lean 4 proof formalization and SymPy calculations. CX Codex +1 ★ 3,951repo stars research-codebase research-codebase — a subagent that automatically documents your codebase. It spawns specialized agents in parallel to locate files, analyze code logic, find usage patterns, and synthesize findings into a single technical report. CX Codex ★ 3,951repo stars search-tools search-tools is a Claude Code skill that routes code searches to LEANN, AST-grep, Morph, or Grep based on the task. CX Codex +1 ★ 3,951repo stars modular-arithmetic Handles modular arithmetic problems in graph number theory. Activates when you need the extended Euclidean algorithm, the Chinese remainder theorem, Euler's theorem, quadratic residues, or order of elements. CX Codex +1 ★ 3,951repo stars agentica-spawn agentica-spawn — a skill that spawns multi-agent patterns via Agentica: swarms for parallel research, hierarchies for role-based tasks, generator-critic loops for iterative refinement, and juries for consensus. CX Codex +1 ★ 3,951repo stars

Project sponsors

Companies that keep this analytics open