Specify a webhook-driven task for OpenClaw
Describe event inputs, validation, deduplication, and intended actions before integration work.
Language-agnostic coding principles for high-quality, maintainable code. Apply when generating new code, reviewing existing code, or refactoring. Covers DRY, KISS, YAGNI, Fail Fast, avoiding over-validation, and generalized SOLID principles (SRP, DIP). Automatically detect violations and provide ...
--- name: clean-code-principles-d985942ebe75 description: Language-agnostic coding principles for high-quality, maintainable code. Apply when generating new code, reviewing existing code, or refactoring. Covers DRY, KISS, YAGNI, Fail Fast, avoiding over-validation, and generalized SOLID principles (SRP, DIP). Automatically detect violations and provide corrected code. --- # Clean Code Principles Apply these principles to all code generation, review, and refactoring tasks. ## Core Principles ### DRY (Don't Repeat Yourself) Extract duplicated logic into reusable functions/modules. Violation signs: copy-pasted code, similar patterns in multiple places. ### KISS (Keep It Simple, Stupid) Prefer straightforward solutions. Avoid: clever one-liners that sacrifice readability, unnecessary abstractions, premature optimization. ### YAGNI (You Aren't Gonna Need It) Implement only what's currently needed. Avoid: speculative features, unused parameters "for future use", over-engineered extensibility. ## Practical Principles ### Fail Fast Detect and report errors at the earliest possible point. Validate at system boundaries (public APIs, user input), not deep in internal logic. ### Avoid Over-Validation Validate once at the entry point. Internal functions can trust their callers. Signs of over-validation: - Same null/type checks repeated across call chain - Defensive checks in private/internal methods - Redundant validation after already-validated data ## Design Principles (Generalized) ### Single Responsibility (SRP) Each function/module should have one reason to change. Split when a function does multiple unrelated tasks. ### Dependency Inversion (DIP) Depend on abstractions, not concrete implementations. Pass dependencies as parameters (functions, interfaces) rather than hardcoding them. ### Open-Closed & Interface Segregation (Light) - OCP: Extend behavior without modifying existing code (via composition, callbacks, plugins) - ISP: Keep interfaces/APIs minimal—expose only what clients need ## Workflow **When generating code**: Apply principles from the start. Prefer simple, focused implementations. **When reviewing code**: Identify violations → explain briefly → provide corrected version. **When refactoring**: Prioritize by impact. Fix critical issues (DRY, over-validation) before stylistic ones. ## Violation Response Format When violations are found, respond with: 1. Brief identification of the issue and which principle it violates 2. Corrected code Keep explanations concise. For detailed examples of each principle, see [references/examples.md](references/examples.md).--- name: clean-code-principles-d985942ebe75 description: Language-agnostic coding principles for high-quality, maintainable code. Apply when generating new code, reviewing existing code, or refactoring. Covers DRY, KISS, YAGNI, Fail Fast, avoiding over-validation, and generalized SOLID principles (SRP, DIP). Automatically detect violations and provide corrected code. --- # Clean Code Principles Apply these principles to all code generation, review, and refactoring tasks. ## Core Principles ### DRY (Don't Repeat Yourself) Extract duplicated logic into reusable functions/modules. Violation signs: copy-pasted code, similar patterns in multiple places. ### KISS (Keep It Simple, Stupid) Prefer straightforward solutions. Avoid: clever one-liners that sacrifice readability, unnecessary abstractions, premature optimization. ### YAGNI (You Aren't Gonna Need It) Implement only what's currently needed. Avoid: speculative features, unused parameters "for future use", over-engineered extensibility. ## Practical Principles ### Fail Fast Detect and report errors at the earliest possible point. Validate at system boundaries (public APIs, user input), not deep in internal logic. ### Avoid Over-Validation Validate once at the entry point. Internal functions can trust their callers. Signs of over-validation: - Same null/type checks repeated across call chain - Defensive checks in private/internal methods - Redundant validation after already-validated data ## Design Principles (Generalized) ### Single Responsibility (SRP) Each function/module should have one reason to change. Split when a function does multiple unrelated tasks. ### Dependency Inversion (DIP) Depend on abstractions, not concrete implementations. Pass dependencies as parameters (functions, interfaces) rather than hardcoding them. ### Open-Closed & Interface Segregation (Light) - OCP: Extend behavior without modifying existing code (via composition, callbacks, plugins) - ISP: Keep interfaces/APIs minimal—expose only what clients need ## Workflow **When generating code**: Apply principles from the start. Prefer simple, focused implementations. **When reviewing code**: Identify violations → explain briefly → provide corrected version. **When refactoring**: Prioritize by impact. Fix critical issues (DRY, over-validation) before stylistic ones. ## Violation Response Format When violations are found, respond with: 1. Brief identification of the issue and which principle it violates 2. Corrected code Keep explanations concise. For detailed examples of each principle, see [references/examples.md](references/examples.md).
Imported from the listed AgentSkill.sh category. Complete supplied package files and notices are retained. Required external tools and credentials are not installed. No bundled code was executed during import. Individual agent task behavior has not been tested.