How tw93/waza works
Waza is a collection of engineering playbooks and helper scripts invoked by coding agents. A user selects a skill for planning, UI work, review, debugging, writing, research, source intake, or agent-setup audits; the skill guides work toward its outcome. README documents the skill capabilities and the general presence of helper scripts, but most skill documents are not in the supplied tree, so their internal orchestration is not mapped. The listed scripts are shown as capability-specific helpers, not as a presumed pipeline.
Main entry points: health/SKILL.md Health audits · read/SKILL.md Read sources · check/SKILL.md Check review · write/SKILL.md Write prose
Components 22
Engineering playbooks
- Think planning
skills/think/SKILL.md - UI building
skills/ui/SKILL.md - Hunt debugging
skills/hunt/SKILL.md - Learn research
skills/learn/SKILL.md
Review and release
- Check review
skills/check/SKILL.md - Audit signals
skills/check/scripts/audit_signals.py - Incident ledger
skills/check/scripts/incident_ledger.py - Release gate
skills/check/scripts/release_gate.py
Agent health
- Health audits
skills/health/SKILL.md - Agent context
skills/health/scripts/check_agent_context.py - Maintainability
skills/health/scripts/check_maintainability.py - Conversation audit
skills/health/scripts/conversation_audit.py - Skill security
skills/health/scripts/scan_skill_security.py - Shell pipe guard
skills/health/scripts/block-pipe-to-shell.py
Read and writing tools
- Read sources
skills/read/SKILL.md - Local fetch
skills/read/scripts/fetch_local.py - Feishu fetch
skills/read/scripts/fetch_feishu.py - Weixin fetch
skills/read/scripts/fetch_weixin.py - DeepSeek fetch
skills/read/scripts/deepseek.py - Write prose
skills/write/SKILL.md - Punctuation check
skills/write/scripts/check_punctuation.py
Other components
- Engineer
Connections 21
- Engineer → Think planning: invokes
README.md - Engineer → UI building: invokes
README.md - Engineer → Check review: invokes
README.md - Engineer → Hunt debugging: invokes
README.md - Engineer → Write prose: invokes
README.md - Engineer → Learn research: invokes
README.md - Engineer → Read sources: invokes
README.md - Engineer → Health audits: invokes
README.md - Check review → Audit signals: may use
- Check review → Incident ledger: may use
- Check review → Release gate: may use
- Health audits → Agent context: may use
- Health audits → Maintainability: may use
- Health audits → Conversation audit: may use
- Health audits → Skill security: may use
- Health audits → Shell pipe guard: may use
- Read sources → Local fetch: may use
- Read sources → Feishu fetch: may use
- Read sources → Weixin fetch: may use
- Read sources → DeepSeek fetch: may use
- Write prose → Punctuation check: may use