Profiles Guide

MCP Profiles Guide

Preset configurations optimized for different use cases: llm-chat, llm-safe, and ci-strict.

What are Profiles?

Profiles are preset configurations that optimize LogicStamp Context for different use cases. Each profile sets specific parameters like dependency depth, code inclusion mode, and maximum nodes per bundle.

Use the profile parameter in logicstamp_refresh_snapshot to select a profile.

llm-chat (Default)

Balanced mode optimized for AI chat interfaces. This is the default profile and recommended for most use cases.

Configuration

  • Depth: 2 (analyzes nested components)
  • Mode: header (contracts + JSDoc headers)
  • Max nodes: 100 per bundle

Best For

  • General AI chat workflows
  • Code review and refactoring
  • Understanding component interfaces
  • Most common use cases

Usage:

logicstamp_refresh_snapshot({ profile: "llm-chat" })

llm-safe

Conservative mode optimized for token-limited contexts. Use this when you need to minimize token usage.

Configuration

  • Depth: 2 (analyzes nested components)
  • Mode: header (contracts + JSDoc headers)
  • Max nodes: 30 per bundle (reduced from 100)

Best For

  • Token-limited AI contexts
  • Large codebases where token count matters
  • When you need maximum token efficiency
  • Cost-sensitive workflows

Usage:

logicstamp_refresh_snapshot({ profile: "llm-safe" })

ci-strict

Strict validation mode optimized for validation workflows. Focuses on contract validation and dependency checking. Useful for CI/CD pipelines (note: git baseline comparison is not yet implemented).

Configuration

  • Depth: 2 (analyzes nested components)
  • Mode: none (contracts only, no source code)
  • Max nodes: 100 per bundle
  • Strict validation: Enabled (fails on missing dependencies)

Best For

  • CI/CD pipelines
  • Contract validation
  • Dependency checking
  • Automated testing

Usage:

logicstamp_refresh_snapshot({ profile: "ci-strict" })

Profile Comparison

ProfileDepthModeMax NodesUse Case
llm-chat1header100General AI workflows (default)
llm-safe1header30Token-limited contexts
ci-strict1none100CI/CD validation

Choosing a Profile

  • Start with llm-chat - This is the default and works well for most use cases
  • Use llm-safe - When you're hitting token limits or need maximum efficiency
  • Use ci-strict - For validation workflows where you need strict dependency checking (note: git baseline comparison for CI/CD is not yet implemented)
  • You can override - Profile settings can be overridden with explicit parameters

Learn More

Explore style metadata, comparison workflows, and best practices.