optimize-integration

Enrich tool descriptions, refine field compaction, and tune responses in Switchboard adapters.

15|7|Updated Feb 26, 2026
One-click install
npx skills add https://github.com/daltoniam/switchboard --skill optimize-integration
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: optimize-integration
Source: https://github.com/daltoniam/switchboard/tree/main/.agents/skills/optimize-integration
Command: npx skills add https://github.com/daltoniam/switchboard --skill optimize-integration

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Improve an existing Switchboard integration adapter's LLM usability by enriching tool descriptions, refining field compaction, and tuning response behavior.

Core Features & Use Cases

  • Phase 1: Tool Description Enrichment — tier and enrich descriptions to guide the LLM's routing decisions.
  • Phase 2: Field Compaction Refinement — tailor output specs to reduce noise while preserving necessary routing information.
  • Phase 3: Response Tuning — calibrate response size and client behavior for raw HTTP adapters and similar.

Quick Start

Run Phase 1 enrichment on the target adapter, then Phase 2 compaction refinement, and finally Phase 3 response tuning.

Frequently Asked Questions about optimize-integration

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I improve LLM routing accuracy for an existing integration adapter?

Improve LLM routing accuracy by enriching tool descriptions, refining field compaction, and tuning response behavior. This optimization process applies iteratively to existing adapters to reduce payload noise and guide routing decisions effectively.

What does field compaction refinement do for an adapter's output spec?

Field compaction refinement tailors output specifications to reduce noise while preserving necessary routing information. This process minimizes irrelevant payload data sent to the LLM, directly improving response efficiency and routing accuracy across the adapter.

How do I calibrate response size and client behavior for raw HTTP adapters?

Calibrate response size and client behavior for raw HTTP adapters through response tuning. This optimization phase adjusts payload shape and client behavior to ensure efficient data delivery and prevent oversized responses from degrading LLM usability.

What's the best way to optimize an integration adapter without disrupting existing functionality?

Optimize integration adapters non-disruptively by applying phase-driven changes across description enrichment, compaction refinement, and response tuning. Commit changes iteratively with unit tests to ensure frontmatter metadata consistency and maintain existing functionality.

When do I need to enrich tool descriptions for an integration adapter?

Enrich tool descriptions when an LLM struggles to correctly route requests to the target adapter. Tiering and enriching descriptions guides the LLM's routing decisions, ensuring it selects the correct tool based on the user's intent.

Does adapter optimization require external dependencies to modify payload shape?

Adapter optimization requires no external dependencies to modify payload shape and field lists. The process operates directly on existing adapter configurations to tune response behavior and refine compaction without additional packages.