Build code from GitHub issue specifications. Implements in chunks with tests after each chunk. Uses just commands for quality gates. Creates self-contained modules.
/plugin marketplace add bryonjacob/aug/plugin install aug-dev@augThis skill inherits all available tools. When active, it can use any tool Claude has access to.
You build modules from GitHub issue specifications.
Apply these principles during development:
GitHub issue with structured metadata:
Parse from issue body, not labels.
Branch: epic/{epic-id}/{task-slug}
Base: main
Target: main (via PR)
All branches from main. Flat structure only.
Extract:
Update docs with retcon writing (present tense):
just format
git commit -m "docs: Update for #{ISSUE}"
git push
For each chunk in issue:
A. Implement
B. Test
C. Verify
just format # Auto-fix
just lint # Auto-fix
just typecheck # Check types
just test # Run tests
D. Commit
git commit -m "feat: Implement {CHUNK} for #{ISSUE}"
git push
Incremental commits. Incremental pushes. Recoverable.
After all chunks:
just coverage # Verify >= 96%
If below threshold:
Add edge cases, error handling, integration tests.
git commit -m "test: Enhance coverage for #{ISSUE}"
git push
just check-all
Run user acceptance tests from issue.
If failures:
just format → just lint → fix types → fix testsjust check-allIf still failing after 3 attempts: escalate (not your job to force it).
Quality gate passed. Ready for PR creation.
module_name/
├── __init__.py # Public interface ONLY
├── README.md # Contract (MANDATORY)
├── core.py # Implementation
├── models.py # Data structures
└── tests/ # Unit tests
# __init__.py
from .core import process
from .models import Input, Output
__all__ = ['process', 'Input', 'Output']
No private exports. No internal leaks.
Before claiming done:
just check-all passingDon't:
Standard interface across all projects:
just format # Auto-fix formatting
just lint # Auto-fix linting
just typecheck # Static type checking
just test # Run test suite
just coverage # Coverage analysis (96% threshold)
just check-all # Full quality gate
Use these. Don't invent custom verification.
Follow:
@ai_context/IMPLEMENTATION_PHILOSOPHY.md@ai_context/MODULAR_DESIGN_PHILOSOPHY.mdCheck: @DISCOVERIES.md for known patterns
When quality checks fail:
Auto-fix what's fixable. Escalate what's not.
Task complete when:
No draft PRs. No "almost done". Either passing or escalated.
This skill should be used when the user asks to "create a slash command", "add a command", "write a custom command", "define command arguments", "use command frontmatter", "organize commands", "create command with file references", "interactive command", "use AskUserQuestion in command", or needs guidance on slash command structure, YAML frontmatter fields, dynamic arguments, bash execution in commands, user interaction patterns, or command development best practices for Claude Code.
This skill should be used when the user asks to "create an agent", "add an agent", "write a subagent", "agent frontmatter", "when to use description", "agent examples", "agent tools", "agent colors", "autonomous agent", or needs guidance on agent structure, system prompts, triggering conditions, or agent development best practices for Claude Code plugins.
This skill should be used when the user asks to "create a hook", "add a PreToolUse/PostToolUse/Stop hook", "validate tool use", "implement prompt-based hooks", "use ${CLAUDE_PLUGIN_ROOT}", "set up event-driven automation", "block dangerous commands", or mentions hook events (PreToolUse, PostToolUse, Stop, SubagentStop, SessionStart, SessionEnd, UserPromptSubmit, PreCompact, Notification). Provides comprehensive guidance for creating and implementing Claude Code plugin hooks with focus on advanced prompt-based hooks API.