菜单
Design tools that agents can use effectively. Use when creating new tools for agents, debugging tool-related failures, or optimizing existing tool sets.
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
基于 SOC 职业分类
| name | tool-design |
| description | Design tools that agents can use effectively. Use when creating new tools for agents, debugging tool-related failures, or optimizing existing tool sets. |
Tools are the primary mechanism through which agents interact with the world. They define the contract between deterministic systems and non-deterministic agents. Unlike traditional software APIs designed for developers, tool APIs must be designed for language models that reason about intent, infer parameter values, and generate calls from natural language requests. Poor tool design creates failure modes that no amount of prompt engineering can fix. Effective tool design follows specific principles that account for how agents perceive and use tools.
Activate this skill when:
Tools are contracts between deterministic systems and non-deterministic agents. The consolidation principle states that if a human engineer cannot definitively say which tool should be used in a given situation, an agent cannot be expected to do better. Effective tool descriptions are prompt engineering that shapes agent behavior.
Key principles include: clear descriptions that answer what, when, and what returns; response formats that balance completeness and token efficiency; error messages that enable recovery; and consistent conventions that reduce cognitive load.
Tools as Contracts Tools are contracts between deterministic systems and non-deterministic agents. When humans call APIs, they understand the contract and make appropriate requests. Agents must infer the contract from descriptions and generate calls that match expected formats.
This fundamental difference requires rethinking API design. The contract must be unambiguous, examples must illustrate expected patterns, and error messages must guide correction. Every ambiguity in tool definitions becomes a potential failure mode.
Tool Description as Prompt Tool descriptions are loaded into agent context and collectively steer behavior. The descriptions are not just documentation—they are prompt engineering that shapes how agents reason about tool use.
Poor descriptions like "Search the database" with cryptic parameter names force agents to guess. Optimized descriptions include usage context, examples, and defaults. The description answers: what the tool does, when to use it, and what it produces.
Namespacing and Organization As tool collections grow, organization becomes critical. Namespacing groups related tools under common prefixes, helping agents select appropriate tools at the right time.
Namespacing creates clear boundaries between functionality. When an agent needs database information, it routes to the database namespace. When it needs web search, it routes to web namespace.
Single Comprehensive Tools The consolidation principle states that if a human engineer cannot definitively say which tool should be used in a given situation, an agent cannot be expected to do better. This leads to a preference for single comprehensive tools over multiple narrow tools.
Instead of implementing list_users, list_events, and create_event, implement schedule_event that finds availability and schedules. The comprehensive tool handles the full workflow internally rather than requiring agents to chain multiple calls.
Why Consolidation Works Agents have limited context and attention. Each tool in the collection competes for attention in the tool selection phase. Each tool adds description tokens that consume context budget. Overlapping functionality creates ambiguity about which tool to use.
Consolidation reduces token consumption by eliminating redundant descriptions. It eliminates ambiguity by having one tool cover each workflow. It reduces tool selection complexity by shrinking the effective tool set.
When Not to Consolidate Consolidation is not universally correct. Tools with fundamentally different behaviors should remain separate. Tools used in different contexts benefit from separation. Tools that might be called independently should not be artificially bundled.
Description Structure Effective tool descriptions answer four questions:
What does the tool do? Clear, specific description of functionality. Avoid vague language like "helps with" or "can be used for." State exactly what the tool accomplishes.
When should it be used? Specific triggers and contexts. Include both direct triggers ("User asks about pricing") and indirect signals ("Need current market rates").
What inputs does it accept? Parameter descriptions with types, constraints, and defaults. Explain what each parameter controls.
What does it return? Output format and structure. Include examples of successful responses and error conditions.
Default Parameter Selection Defaults should reflect common use cases. They reduce agent burden by eliminating unnecessary parameter specification. They prevent errors from omitted parameters.
Tool response size significantly impacts context usage. Implementing response format options gives agents control over verbosity.
Concise format returns essential fields only, appropriate for confirmation or basic information. Detailed format returns complete objects with all fields, appropriate when full context is needed for decisions.
Include guidance in tool descriptions about when to use each format. Agents learn to select appropriate formats based on task requirements.
Error messages serve two audiences: developers debugging issues and agents recovering from failures. For agents, error messages must be actionable. They must tell the agent what went wrong and how to correct it.
Design error messages that enable recovery. For retryable errors, include retry guidance. For input errors, include corrected format. For missing data, include what's needed.
Use a consistent schema across all tools. Establish naming conventions: verb-noun pattern for tool names, consistent parameter names across tools, consistent return field names.
Research shows tool description overlap causes model confusion. More tools do not always lead to better outcomes. A reasonable guideline is 10-20 tools for most applications. If more are needed, use namespacing to create logical groupings.
Implement mechanisms to help agents select the right tool: tool grouping, example-based selection, and hierarchy with umbrella tools that route to specialized sub-tools.
When using MCP (Model Context Protocol) tools, always use fully qualified tool names to avoid "tool not found" errors.
Format: ServerName:tool_name
# Correct: Fully qualified names
"Use the BigQuery:bigquery_schema tool to retrieve table schemas."
"Use the GitHub:create_issue tool to create issues."
# Incorrect: Unqualified names
"Use the bigquery_schema tool..." # May fail with multiple servers
Without the server prefix, agents may fail to locate tools, especially when multiple MCP servers are available. Establish naming conventions that include server context in all tool references.
Claude can optimize its own tools. When given a tool and observed failure modes, it diagnoses issues and suggests improvements. Production testing shows this approach achieves 40% reduction in task completion time by helping future agents avoid mistakes.
The Tool-Testing Agent Pattern:
def optimize_tool_description(tool_spec, failure_examples):
"""
Use an agent to analyze tool failures and improve descriptions.
Process:
1. Agent attempts to use tool across diverse tasks
2. Collect failure modes and friction points
3. Agent analyzes failures and proposes improvements
4. Test improved descriptions against same tasks
"""
prompt = f"""
Analyze this tool specification and the observed failures.
Tool: {tool_spec}
Failures observed:
{failure_examples}
Identify:
1. Why agents are failing with this tool
2. What information is missing from the description
3. What ambiguities cause incorrect usage
Propose an improved tool description that addresses these issues.
"""
return get_agent_response(prompt)
This creates a feedback loop: agents using tools generate failure data, which agents then use to improve tool descriptions, which reduces future failures.
Evaluate tool designs against criteria: unambiguity, completeness, recoverability, efficiency, and consistency. Test tools by presenting representative agent requests and evaluating the resulting tool calls.
Vague descriptions: "Search the database for customer information" leaves too many questions unanswered.
Cryptic parameter names: Parameters named x, val, or param1 force agents to guess meaning.
Missing error handling: Tools that fail with generic errors provide no recovery guidance.
Inconsistent naming: Using id in some tools, identifier in others, and customer_id in some creates confusion.
When designing tool collections:
Example 1: Well-Designed Tool
def get_customer(customer_id: str, format: str = "concise"):
"""
Retrieve customer information by ID.
Use when:
- User asks about specific customer details
- Need customer context for decision-making
- Verifying customer identity
Args:
customer_id: Format "CUST-######" (e.g., "CUST-000001")
format: "concise" for key fields, "detailed" for complete record
Returns:
Customer object with requested fields
Errors:
NOT_FOUND: Customer ID not found
INVALID_FORMAT: ID must match CUST-###### pattern
"""
Example 2: Poor Tool Design
This example demonstrates several tool design anti-patterns:
def search(query):
"""Search the database."""
pass
Problems with this design:
Failure modes:
This skill connects to:
Internal reference:
Related skills in this collection:
External resources:
Created: 2025-12-20 Last Updated: 2025-12-20 Author: Agent Skills for Context Engineering Contributors Version: 1.0.0
Set up a FREE local LLM (via Ollama) and connect Claude Code to it so it runs offline with no API costs. Use this whenever the user wants to run Claude Code on a local/self-hosted model, use Ollama/Qwen/Llama/Mistral/DeepSeek with Claude Code, get a "free" or "offline" coding agent, avoid Anthropic API costs, point ANTHROPIC_BASE_URL at localhost, or asks "can my machine/VPS run a local LLM for Claude Code". Trigger even if they only say "run Claude Code locally", "local model for coding", "self-hosted Claude Code", or name a local model (qwen3, llama3, deepseek-coder). Handles install, model tuning, wiring, and an honest hardware reality-check (a capable model AND fast hardware are BOTH required).
Production-grade Docker containerization for Python and Node.js applications. This skill should be used when users ask to containerize applications, create Dockerfiles, dockerize projects, or set up Docker Compose. Auto-detects project structure, analyzes .env for secrets, validates security, and generates tested Dockerfiles.
Production-grade Helm 4 chart development, release management, and debugging. This skill should be used when users ask to create Helm charts, deploy with Helm, manage releases (install/upgrade/rollback), push charts to OCI registries, debug failed deployments, configure chart dependencies, create umbrella charts, set up GitOps with ArgoCD/Flux, or troubleshoot Helm issues. Auto-detects from Dockerfile/code, generates production-hardened charts with library patterns. Complements kubernetes skill.
Apache Kafka on Kubernetes with Strimzi (KRaft mode, no ZooKeeper). This skill should be used when users ask to deploy Kafka clusters, build producers/consumers, implement event-driven patterns, or debug Kafka issues. Includes tested manifests and Makefile for one-command deployment.
Production-grade Kubernetes manifests and debugging for containerized applications. This skill should be used when users ask to deploy to Kubernetes, create K8s manifests, containerize for K8s, set up Deployments/Services/Jobs/StatefulSets/CronJobs, create namespaces with resource quotas, set up multi-team isolation, configure ResourceQuota/ LimitRange, secure with RBAC (ServiceAccount, Role, RoleBinding), configure init containers (model download, db wait, migrations), set up sidecars (logging, metrics), or debug pods (CrashLoopBackOff, logs, exec, describe, events). Auto-detects from Dockerfile/code, generates hardened manifests with educational comments. CKAD-aligned.
Build evaluation frameworks for agent systems. Use when testing agent performance, validating context engineering choices, or measuring improvements over time.