autonomous-development_skill

Overview

This skill offers end-to-end autonomous development strategies for managing the full software lifecycle with minimal human intervention. It covers milestone planning, incremental implementation, automated debugging, parameter validation, and continuous quality assurance. The goal is production-ready outputs with reproducible, privacy-first processes.

How this skill works

The skill decomposes high-level requirements into sequenced milestones using a complexity matrix and dependency-first ordering. It enforces a commit-per-milestone workflow, progressive enhancement phases, and automated validation loops that detect and auto-fix common errors (syntax, imports, types) while escalating complex issues for review. Quality is measured with configurable scorecards and prioritized auto-fix rules.

When to use it

• Implementing new features from high-level requirements
• Managing multi-phase or cross-team development projects
• Running autonomous agents that must maintain code quality and safety
• Automating debugging and repeatable fix loops for CI failures
• Validating cross-file parameter contracts and API contracts

Best practices

• Decompose each requirement into small, measurable milestones with clear complexity estimates
• Commit only working states per milestone and use conventional commit messages
• Follow progressive enhancement: implement a simple proof, then add robustness, optimization, and polish
• Automate common fixes first (syntax, imports, type mismatches) and reserve human review for architectural changes
• Enforce test-first generation for critical paths and maintain quality thresholds before release

Example use cases

• Autonomously implement a new backend feature: dependency setup, core logic, integration tests, docs
• Auto-debug CI failures with up to five iterative repair attempts before requesting human intervention
• Validate parameter consistency across modules and config files before integration tests run
• Run quality scoring and automated fixes to reach a predefined release threshold (e.g., ≥85)
• Generate prioritized test suites and enforce test-first development for core workflows