OpenSkillIndex← back to search
claude-codedevelopment

Babysitter implementation-execution

Execute development tasks to build features, producing code, tests, and configuration artifacts that satisfy specification requirements and comply with constitution standards.

install
source · Clone the upstream repo
git clone https://github.com/a5c-ai/babysitter
Claude Code · Install into ~/.claude/skills/
T=$(mktemp -d) && git clone --depth=1 https://github.com/a5c-ai/babysitter "$T" && mkdir -p ~/.claude/skills && cp -r "$T/library/methodologies/spec-kit/skills/implementation-execution" ~/.claude/skills/a5c-ai-babysitter-implementation-execution && rm -rf "$T"
manifest: library/methodologies/spec-kit/skills/implementation-execution/SKILL.md
tags
#code-generation#test-creation#config-artifact#spec-compliance#task-execution
source content

Implementation Execution

Overview

Execute development tasks from the task list against the specification and constitution. Produces code, tests, and configuration artifacts with progressive validation. Independent tasks execute in parallel; dependent tasks execute sequentially respecting the dependency graph.

When to Use

  • After task list is approved and analysis confirms readiness
  • When implementing feature code against a specification
  • When running per-task tests during implementation
  • When validating implementation against specification requirements

Key Principle

Every line of code must trace back to a specification requirement. Code quality must meet constitution standards. Tests must verify acceptance criteria from the task definition.

Process

  1. Prepare environment - Scaffold directories and configuration
  2. Execute independent tasks - Parallel execution for tasks without dependencies
  3. Execute dependent tasks - Sequential execution respecting dependency graph
  4. Run per-task tests - Verify each task's acceptance criteria
  5. Validate against spec - Check requirement coverage and story satisfaction
  6. Run integration tests - Cross-component testing
  7. Human review - Approve implementation before quality checklist

Tool Use

Invoke via babysitter process: 

methodologies/spec-kit/spec-kit-implementation
Full pipeline: 
methodologies/spec-kit/spec-kit-orchestrator