BMAD-METHOD/bmad-core/checklists/session-kickoff-checklist.md

Session Kickoff Checklist

This checklist ensures AI agents have complete project context and understanding before starting work. It provides systematic session initialization across all agent types.

[[LLM: INITIALIZATION INSTRUCTIONS - SESSION KICKOFF

This is the FIRST checklist to run when starting any new AI agent session. It prevents context gaps, reduces mistakes, and ensures efficient work.

IMPORTANT: This checklist is mandatory for:

• New AI sessions on existing projects
• After significant time gaps (>24 hours)
• When switching between major project areas
• After major changes or pivots
• When onboarding new team members

The goal is to establish complete context BEFORE any work begins.]]

1. MEMORY BANK REVIEW

LLM: Memory Bank is the primary source of project truth. Review systematically, noting dates and potential staleness.

1.1 Core Memory Bank Files

• projectbrief.md reviewed - Project foundation, goals, and scope understood
• activeContext.md reviewed - Current priorities and immediate work identified
• progress.md reviewed - Project state and completed features understood
• systemPatterns.md reviewed - Architecture patterns and decisions noted
• techContext.md reviewed - Technology stack and constraints clear
• productContext.md reviewed - Problem space and user needs understood
• Last update timestamps noted for each file
• Potential inconsistencies between files identified

1.2 Memory Bank Health Assessment

• Files exist and are accessible
• Information appears current (updated within last sprint)
• No major gaps in documentation identified
• Cross-references between files are consistent
• Action items for updates noted if needed

1.3 Project Structure Verification

LLM: Reference project-scaffolding-preference.md for standard project structure. Verify actual structure aligns with BMAD conventions.

• Project follows standard directory structure
• BMAD-specific directories exist (docs/memory-bank, docs/adr, docs/devJournal)
• Documentation directories properly organized
• Source code organization follows conventions
• Test structure aligns with project type

2. ARCHITECTURE DOCUMENTATION

LLM: Architecture drives implementation. Understand the system design thoroughly.

2.1 Architecture Documents

• Primary architecture document located and reviewed
• Document type identified (greenfield, brownfield, frontend, fullstack)
• Core architectural decisions understood
• System components and relationships clear
• Technology choices and versions noted
• API documentation reviewed if exists
• Database schemas understood if applicable

2.2 Architecture Alignment

• Architecture aligns with Memory Bank information
• Recent changes or updates identified
• ADRs reviewed for architectural decisions
• Integration points clearly understood
• Deployment architecture reviewed

3. DEVELOPMENT HISTORY

LLM: Recent history provides context for current work and challenges.

3.1 Dev Journal Review

• Located Dev Journal entries (last 3-5)
• Recent work and decisions understood
• Challenges and blockers identified
• Technical debt or issues noted
• Patterns in development identified
• Key learnings extracted

3.2 ADR Review

• Recent ADRs reviewed (last 3-5)
• Current architectural decisions understood
• Superseded decisions noted
• Pending decisions identified
• ADR alignment with architecture verified

4. CURRENT PROJECT STATE

LLM: Understanding the current state prevents duplicate work and conflicts.

4.1 Git Status Check

• Current branch identified
• Clean working directory confirmed
• Recent commits reviewed (last 10)
• Outstanding changes understood
• Merge conflicts checked
• Remote synchronization status

4.2 Project Health

• Build status checked
• Test suite status verified
• Known failing tests documented
• Blocking issues identified
• Dependencies up to date
• Security vulnerabilities checked

5. SPRINT/ITERATION CONTEXT

LLM: Align work with current sprint goals and priorities.

5.1 Sprint Status

• Current sprint identified
• Sprint goals understood
• User stories in progress identified
• Completed stories this sprint noted
• Sprint timeline clear
• Team velocity understood

5.2 Priority Alignment

• Immediate priorities identified
• Blockers and dependencies clear
• Next planned work understood
• Risk areas identified
• Resource constraints noted

6. CONSISTENCY VALIDATION

LLM: Inconsistencies cause confusion and errors. Identify and flag them.

6.1 Cross-Reference Check

• Memory Bank aligns with codebase reality
• Architecture matches implementation
• ADRs reflected in current code
• Dev Journal matches git history
• Documentation current with changes

6.2 Gap Identification

• Missing documentation identified
• Outdated sections flagged
• Undocumented decisions noted
• Knowledge gaps listed
• Update requirements documented

7. AGENT-SPECIFIC CONTEXT

LLM: Different agents need different context emphasis.

7.1 Role-Based Focus

For Architect:

• Architectural decisions and rationale clear
• Technical debt understood
• Scalability considerations reviewed
• System boundaries defined

For Developer:

• Current implementation tasks clear
• Coding patterns understood
• Testing requirements known
• Local setup verified

For PM/PO:

• Requirements alignment verified
• User stories prioritized
• Stakeholder needs understood
• Timeline constraints clear

For QA:

• Test coverage understood
• Quality gates defined
• Known issues documented
• Testing strategy clear

7.2 Handoff Context

• Previous agent's work understood
• Pending decisions identified
• Open questions documented
• Next steps clear

8. RECOMMENDED ACTIONS

LLM: Based on the review, what should happen next?

8.1 Immediate Actions

• Most urgent task identified
• Blockers that need resolution listed
• Quick wins available noted
• Risk mitigation needed specified

8.2 Documentation Updates

• Memory Bank updates needed listed
• Architecture updates required noted
• ADRs to be created identified
• Dev Journal entries planned

8.3 Strategic Considerations

• Technical debt to address
• Architectural improvements needed
• Process improvements suggested
• Knowledge gaps to fill

SESSION KICKOFF SUMMARY

[[LLM: Generate a concise summary report with:

1. Project Context

   • Project name and purpose
   • Current phase/sprint
   • Key technologies

2. Documentation Health

   • Memory Bank status (Current/Outdated/Missing)
   • Architecture status
   • Overall documentation quality

3. Current State

   • Active work items
   • Recent completions
   • Immediate blockers

4. Inconsistencies Found

   • List any misalignments
   • Documentation gaps
   • Update requirements

5. Recommended Next Steps

   • Priority order
   • Estimated effort
   • Dependencies

Keep it action-oriented and concise.]]

Summary Report

Status: [Complete/Partial/Blocked]

Key Findings:

• Documentation Health: [Good/Fair/Poor]
• Project State: [On Track/At Risk/Blocked]
• Context Quality: [Complete/Adequate/Insufficient]

Priority Actions:

1. [Most urgent action]
2. [Second priority]
3. [Third priority]

Blockers:

• [List any blocking issues]

Agent Ready: [Yes/No - with reason if No]