Implement a spec document phase-by-phase, writing robust idiomatic code that follows codebase patterns.
Implement the given spec document. Work phase-by-phase, writing robust code that follows codebase patterns, and industry best practices and patterns.
Process
- Read the spec - Understand phases, success criteria, and scope boundaries
- Explore the codebase - Read files mentioned in the spec and related code
- Analyze patterns - Identify existing conventions, architecture, and idioms to follow
- Implement phase-by-phase - Complete each phase fully before proceeding
- Verify your work - Run
make checkafter code changes (skip for docs-only)
Implementation Rules
- Follow the plan - The spec is your contract; implement what's specified
- Match codebase patterns - Use existing conventions, not new ones
- Write robust code - Handle errors, edge cases, and failure modes
- Be idiomatic - Use language best practices and established patterns
- No shortcuts - Implement fully, don't stub or placeholder
Per-Phase Workflow
For each phase:
- Read - Understand what the phase requires
- Explore - Read existing code that will be modified or extended
- Implement - Write code matching codebase style and patterns
- Test - Run
make checkto verify (for code changes) - Update - Check off completed items in the spec file
Verification
- Code changes: Run
make checkbefore proceeding to next phase - On failure: Run
make fixfirst, then re-runmake check - Documentation-only: Skip verification
Communication
- If the plan doesn't match codebase reality, explain the discrepancy
- If you need to deviate, explain why before making changes
- Update checkboxes in the spec as you complete each section
You Might Also Like
Related Skills

coding-agent
Run Codex CLI, Claude Code, OpenCode, or Pi Coding Agent via background process for programmatic control.
openclaw
add-uint-support
Add unsigned integer (uint) type support to PyTorch operators by updating AT_DISPATCH macros. Use when adding support for uint16, uint32, uint64 types to operators, kernels, or when user mentions enabling unsigned types, barebones unsigned types, or uint support.
pytorch
at-dispatch-v2
Convert PyTorch AT_DISPATCH macros to AT_DISPATCH_V2 format in ATen C++ code. Use when porting AT_DISPATCH_ALL_TYPES_AND*, AT_DISPATCH_FLOATING_TYPES*, or other dispatch macros to the new v2 API. For ATen kernel files, CUDA kernels, and native operator implementations.
pytorch
skill-writer
Guide users through creating Agent Skills for Claude Code. Use when the user wants to create, write, author, or design a new Skill, or needs help with SKILL.md files, frontmatter, or skill structure.
pytorch
implementing-jsc-classes-cpp
Implements JavaScript classes in C++ using JavaScriptCore. Use when creating new JS classes with C++ bindings, prototypes, or constructors.
oven-sh
implementing-jsc-classes-zig
Creates JavaScript classes using Bun's Zig bindings generator (.classes.ts). Use when implementing new JS APIs in Zig with JSC integration.
oven-sh