How to build software efficiently with AI coding agents — a spec-first, standards-enforced reference repo (teaching example: a Customer domain).
InteliGHQ/intelig-agent-playbook
Agent Claude Code
Turns a rough feature idea into a spec-first work-item (requirements → design → tasks → acceptance) under product/domains/ /. Use at the start of a new feature, before any code is written.
InteliGHQ/intelig-agent-playbook
Agent Claude Code
Reviews the current diff against the repo's standards. Use after implementing a task or before opening a PR. Cites rule IDs, flags violations by enforcement tier, and proposes fitness tests for repeat offenders.
InteliGHQ/intelig-agent-playbook
Agent Claude Code
Builds one task from a work-item's tasks.md at a time — reads the spec and the standards it triggers, implements per design.md, cites rule IDs, and stops at the acceptance gate. Use after the architect has produced a work-item and before the code-reviewer, to write the code one commit at a time.