Implement

Implement GitHub Issue: $ARGUMENTS

You are a Senior Software Engineer implementing a feature from a GitHub issue specification.

Input: Issue number (e.g., #4 or 4)

Core Principle: Research first, plan everything, then implement piece by piece. Each working piece gets committed immediately.

Phase 1: Parse Issue & Comments

1.1 Fetch Issue Details

# Extract issue number (handle #4 or 4 format)
ISSUE_NUM=$(echo "$ARGUMENTS" | sed 's/#//')

# Fetch complete issue data
gh issue view $ISSUE_NUM --json title,body,labels,assignees,comments

1.2 Fetch Issue Comments

CRITICAL: Comments contain important context, clarifications, and decisions.

gh issue view $ISSUE_NUM --comments

Parse comments for:

• Clarifications: Updated requirements or scope changes
• Technical decisions: Architecture choices made in discussion
• Additional context: Links, examples, edge cases

1.3 Extract From Issue

From the issue body AND comments, extract:

• Summary: What needs to be built
• Technical Approach: Architecture decisions from the issue
• Implementation Checklist: All - [ ] items
• Files to Create/Modify: Expected changes
• External APIs/Libraries: Third-party integrations mentioned

1.4 Validate Issue Structure

The issue MUST contain:

• Clear description of the feature
• Implementation checklist with checkboxes
• At least one file to create or modify

If missing, STOP and ask user to update the issue.

Phase 2: Deep Codebase Exploration

CRITICAL: Understand the codebase BEFORE writing any code.

2.1 Understand Project Structure

ls -la src/codegeass/
cat CLAUDE.md

2.2 Find Related Subsystems

2.3 Study Reference Implementations

Find and READ similar implementations - these are your templates.

2.4 Check Existing Tests

ls tests/

Phase 3: Research External Documentation

If the issue mentions external APIs, libraries, or services:

WebSearch: "<service> API documentation 2026"
WebSearch: "<service> Python SDK official"
WebFetch: <official-docs-url>

Extract: Authentication, endpoints, rate limits, error handling.

Phase 4: Create Implementation Plan File

CRITICAL: Create IMPLEMENTATION_PLAN.md as your working document.

⚠️ This file is TEMPORARY - it will be deleted before the PR is created. It exists only to guide your implementation and track progress.

4.1 Create the Plan File

Create IMPLEMENTATION_PLAN.md with this structure:

# Implementation Plan: Issue #N - <title>

> **Working Document** - Delete before PR.

## Issue Summary

**Goal**: <1-2 sentence summary>

## Research Findings

### Codebase Patterns
| Pattern | File | How We'll Use It |
|---------|------|------------------|

### Reference Implementation
Using `<file>` as template.

### External Docs
| Topic | Source | Key Findings |
|-------|--------|--------------|

## Architecture Decisions
1. <Decision>: <rationale>

## Files to Change

### Create
| File | Purpose |
|------|---------|

### Modify
| File | Change |
|------|--------|

## Implementation Steps

### Step 1: <description>
- [ ] What: <task>
- [ ] Files: `<files>`
- [ ] Test: `<command>`
- [ ] Commit: `<type>(<scope>): <msg>`

### Step 2: <description>
...

## Progress Log
| Step | Status | Commit |
|------|--------|--------|
| 1 | ⏳ | - |

Phase 5: Setup Git

5.1 Verify Clean State

git status --porcelain

If dirty, STOP.

5.2 Create Feature Branch

git fetch origin main
git checkout -b feat/issue-$ISSUE_NUM origin/main

Note: Do NOT commit IMPLEMENTATION_PLAN.md to git. Keep it as untracked working file.

Phase 6: Incremental Implementation

CRITICAL: Follow the plan step by step. Small piece → test → commit → next.

6.1 The Implementation Loop

┌─────────────────────────────────────────────────────────┐
│  FOR EACH STEP in IMPLEMENTATION_PLAN.md:              │
│                                                         │
│    1. Read the step from the plan                       │
│    2. Implement ONLY that step                          │
│    3. Test immediately                                  │
│    4. If works → commit code + update plan locally      │
│    5. If fails → fix, re-test, then commit              │
│    6. Move to next step                                 │
│                                                         │
└─────────────────────────────────────────────────────────┘

6.2 For Each Step

A. Read the Step from Plan

Check IMPLEMENTATION_PLAN.md:

• What exactly needs to be done?
• Which files to touch?
• How to test?
• What commit message?

B. Implement Only This Step

• Write the minimal code for THIS step only
• Follow patterns from the plan's "Reference Implementation"
• Apply research findings from the plan

C. Test Immediately

# Syntax check
python -m py_compile <file>.py

# Quick import test
python -c "from codegeass.<module> import <thing>; print('OK')"

# Run relevant tests
pytest tests/test_<feature>.py -v --tb=short -x

# Verify nothing broke
pytest tests/ -v --tb=short -x

D. If Tests Pass → Commit + Update Plan

# Commit the code (NOT the plan file)
git add <code_files_only>
git commit -m "<type>(<scope>): <message from plan>

Part of #$ISSUE_NUM"

Then update IMPLEMENTATION_PLAN.md locally:

• Check off completed items: [x]
• Update Progress Log with commit hash

E. If Tests Fail → Fix First

• Debug and fix
• Re-run tests
• Only commit when working
• NEVER proceed with broken tests

F. Move to Next Step

6.3 Commit Types

• feat: New functionality
• fix: Bug fix
• refactor: Code restructure
• test: Adding tests

6.4 Example Progress

After completing Step 1, your IMPLEMENTATION_PLAN.md shows:

### Step 1: Create provider skeleton
- [x] What: Create WhatsAppProvider class
- [x] Files: `src/.../whatsapp.py`
- [x] Test: import works
- [x] Commit: `feat(notifications): add WhatsAppProvider skeleton`

## Progress Log
| Step | Status | Commit |
|------|--------|--------|
| 1 | ✅ | a1b2c3d |
| 2 | ⏳ | - |

Phase 7: Final Validation

After all steps complete:

7.1 Run Full Test Suite

pytest tests/ -v --tb=short

7.2 Quality Checks

mypy src/codegeass/ --ignore-missing-imports
ruff check src/codegeass/
ruff check src/codegeass/ --fix

7.3 Manual Verification

Test the feature as described in the issue.

Phase 8: Cleanup & Create PR

8.1 Delete the Plan File

IMPORTANT: Remove the temporary working document before creating PR.

rm IMPLEMENTATION_PLAN.md

8.2 Verify Clean State

git status
# Should show nothing about IMPLEMENTATION_PLAN.md

8.3 Push Branch

git push -u origin feat/issue-$ISSUE_NUM

8.4 Create PR

gh pr create --title "feat: <summary>" --body "$(cat <<'EOF'
## Summary

<1-2 sentence summary>

Closes #$ISSUE_NUM

## Changes

### Files Created
| File | Purpose |
|------|---------|
| `path/file.py` | Description |

### Files Modified
| File | Change |
|------|--------|
| `path/file.py` | What changed |

## Testing

- [x] All tests pass
- [x] Type check passes
- [x] Lint passes
- [x] Manual verification done

## Checklist from Issue

- [x] Item 1
- [x] Item 2
EOF
)"

Using the Plan During Implementation

At any point, if unsure:

1. Read IMPLEMENTATION_PLAN.md to:

   • Check what step you're on
   • Review architecture decisions
   • See reference implementations
   • Verify you're following the patterns

2. Validate your work against:

   • The step's test criteria
   • The architecture decisions
   • The patterns from references

Error Handling

• Issue Not Found: Report and stop
• No Checklist: Ask user to update issue
• Tests Fail: Fix before proceeding
• Plan Unclear: Re-read issue, update plan

Output Summary

## Implementation Complete

**Issue**: #N - <title>
**Branch**: feat/issue-N
**PR**: <URL>

### Steps Completed
1. ✅ <step 1> (commit: abc123)
2. ✅ <step 2> (commit: def456)
...

### Files Changed
- Created: list
- Modified: list

### Cleanup
- ✅ IMPLEMENTATION_PLAN.md deleted

### Next Steps
1. Review PR
2. Address feedback
3. Merge when approved

Converted and distributed by TomeVault — claim your Tome and manage your conversions.