coding
General-purpose software engineering assistant: implement features, write functions, scaffold projects, and produce idiomatic code in many languages.
SKILL.mdskills/development/coding/SKILL.md
GitHubCoding
Purpose
Help the user write, modify, and reason about code in any language. Focus on idiomatic, testable, readable code with clear interfaces.
When to Use
- User asks to implement a feature or function
- User wants help understanding or modifying existing code
- User is starting a new project or module
When NOT to Use
- The bug is non-obvious (route to
debugging) - The user wants a security review (route to
code-reviewthensecure-coding) - The user wants architectural design (route to
architecture)
Capabilities
- Implement features and functions
- Refactor for clarity and idiom
- Write idiomatic code for the chosen language
- Suggest tests for new code
- Document public APIs
Inputs
- Language / framework
- Desired behavior
- Constraints (performance, style, dependencies)
Tools
- The language toolchain
- Standard library reference
- Optional: linter, formatter
Examples
User: "Write a Python function to dedupe a list of dicts by 'id'." Response: Idiomatic implementation with type hints, docstring, edge-case handling, and a unit test.
Safety
- Don't write code that accesses credentials without consent
- Don't write code that performs destructive filesystem operations without confirmation
- Cite sources when borrowing non-trivial patterns
Source
Custom skill, written for this library.
Notes
Pairs with:
debugging(when something fails)testing(alongside implementation)code-review(before merging)architecture(for larger decisions)
Workflow
- Confirm the goal, scope, and constraints.
- Inspect the available context before acting.
- Apply the skill-specific guidance in this document.
- Verify the result and report limitations or next steps.