Workflow
Explore the core processes and decision-making flows of SidequestLab
0+
Decisions
0
Enforced Rules
0 Steps
Quality Gates
0 Types
LLM Models
Role Legend
workflow_page.legend.chairman
workflow_page.legend.ceo
workflow_page.legend.dev
workflow_page.legend.qa
workflow_page.legend.devops
workflow_page.legend.advisor
workflow_page.legend.historian
Project Lifecycle
From PRD creation to post-deploy verification, with S/A/B/L grade classification and 3-tier QA verification
Key Constraints
- No production deployment without QA verification
- All projects start after PRD approval
- Post-deploy URL verification required (PROJECTS.md SSOT)
Decision-Making Process
From issue occurrence to rule enforcement, with CEO-Advisor check system and incident-to-rule closure loop
Key Constraints
- CEO cannot use Edit/Write tools
- Board Advisor cross-check mandatory
- All decisions recorded with DEC number
8-Step Quality Gate
Automated 8-step verification pipeline from Syntax to Integration. Dev-QA-DevOps 3-tier role separation
Key Constraints
- Rollback to failed step on failure
- 3-tier: Dev(functional) - QA(quality) - DevOps(deploy)
- 80%+ coverage required
Tech Review Protocol
Company fit pre-check comes first, objectivity ensured through Board Advisor cross-check
Key Constraints
- Company fit check comes first
- Board Advisor cross-check mandatory
- Review results recorded in DECISIONS.md
Harness Engineering
How Do We Ensure AI Agent Quality?
L0 Safety Net, L1 Enablement, L2 Traceable Ops β Harness Engineering System perfected over 6 versions
L0
Safety Net
Automated defense via pre-commit hooks
L1
Enablement
Custom Subagent Architecture
L2
Traceable Ops
Traceable operations via run_id
All these processes are operated by AI agents
Only one human β the Chairman. Everyone else is an AI agent.
About Us