AI section centralizes practical guidance for building repeatable agent workflows in your engineering process.
Why this section exists
Operational consistency
Keep execution style and quality standards stable across tasks and contributors.
Faster delivery
Reduce prompt rewriting and context switching with reusable workflow units.
Lower risk
Encode guardrails for safety, validation, and production-ready changes.
Scalable process
Turn best practices into reusable assets your team can apply repeatedly.
Start with Skills
Skills
Learn how to install, organize, and run Rewrite Skills in real workflows, including
npx skills add.Typical AI workflow lifecycle
1. Define scope
1. Define scope
State the goal, ownership boundaries, and acceptance criteria before execution starts.
2. Execute with a skill
2. Execute with a skill
Use a focused skill to guide discovery, implementation, and validation with less variance.
3. Verify outcomes
3. Verify outcomes
Confirm behavior with tests, docs checks, and regression safeguards before shipping.
4. Iterate the skill
4. Iterate the skill
Update the skill with lessons learned so future runs become faster and safer.
Next step
Open Skills guide
Full reference with installation command, usage patterns, and best practices.
Ecosystem
See packages that support AI-driven implementation workflows.