Brainforge Knowledge

PLAN-OVERVIEW-TEMPLATE

3 min read

Plan Overview: {Feature Name}

This is a decomposed plan overview. It coordinates multiple sub-plans that together implement a complex feature. Each sub-plan is self-contained and executable in a fresh conversation. Do NOT implement from this file — use /execute on each sub-plan in order.

Total Sub-Plans: {N} Total Estimated Tasks: {total across all sub-plans}

Feature Description

{What are we building? One paragraph overview.}

User Story

As a {user type}, I want to {action}, so that {benefit}.

Problem Statement

{Why are we building this? What specific problem or opportunity does it address?}

Solution Statement

{What approach did we choose and why? Capture decisions from vibe planning.}

  • Decision 1: {choice} — because {reason}
  • Decision 2: {choice} — because {reason}

Feature Metadata

  • Feature Type: {New Capability / Enhancement / Refactor / Bug Fix}
  • Estimated Complexity: High
  • Plan Mode: Decomposed ({N} sub-plans)
  • Primary Systems Affected: {list all components/services}
  • Dependencies: {external libraries or services required}

CONTEXT REFERENCES

Shared context that ALL sub-plans need. Each sub-plan also has its own per-phase context section. The execution agent reads BOTH this section and the sub-plan’s context before implementing.

Relevant Codebase Files

IMPORTANT: The execution agent MUST read these files before implementing any sub-plan!

  • path/to/file (lines X-Y) — Why: {contains pattern for Z that we’ll mirror}
  • path/to/file (lines X-Y) — Why: {database model structure to follow}

New Files to Create (All Sub-Plans)

  • path/to/new_file — {purpose} — Sub-plan: {NN}
  • path/to/new_file — {purpose} — Sub-plan: {NN}
  • Memory: {summary} — Relevance: {why this matters}
  • (If no relevant memories found, write “No relevant memories found in memory.md”)

Relevant Documentation

Patterns to Follow

{Pattern Name} (from path/to/file:lines):

{actual code snippet from the project}
  • Why this pattern: {explanation}
  • Common gotchas: {warnings}

PLAN INDEX

#PhaseSub-Plan FileTasksContext Load
01{phase-name}requests/{feature}-plan-01-{phase}.md{N}{Low/Medium}
02{phase-name}requests/{feature}-plan-02-{phase}.md{N}{Low/Medium}
03{phase-name}requests/{feature}-plan-03-{phase}.md{N}{Low/Medium}

Each sub-plan targets 5-8 tasks and 500-700 lines. Sub-plans must be self-contained — the execution agent has no memory of previous sub-plans. Include full context references, pattern examples, and detailed task specifications. Context load estimates help decide instance assignment (Low = minimal codebase reads, Medium = several files).

EXECUTION ROUTING

Each sub-plan runs in a fresh /execute session. Recommended: Use Sonnet for execution, Opus for planning.

Execution Instructions

Manual execution (recommended for complex features):

# Execute each sub-plan in a fresh session
claude
> /execute requests/{feature}-plan-01-{phase}.md
 
claude
> /execute requests/{feature}-plan-02-{phase}.md
 
claude
> /execute requests/{feature}-plan-03-{phase}.md

Automated execution (via claude -p):

for plan in requests/{feature}-plan-*.md; do
  claude -p "/execute $plan" --model sonnet
done

Between sub-plans:

  • Each sub-plan runs in a fresh conversation (context reset)
  • Read HANDOFF NOTES from completed sub-plan before starting the next
  • If a sub-plan fails, fix and re-run it before proceeding

ACCEPTANCE CRITERIA

  • {Feature-wide criterion 1}
  • {Feature-wide criterion 2}
  • All sub-plans executed successfully
  • All sub-plan validation commands pass
  • No broken cross-references between sub-plan outputs
  • Backwards compatibility maintained
  • Documentation updated

COMPLETION CHECKLIST

  • Sub-plan 01 ({phase}) — complete
  • Sub-plan 02 ({phase}) — complete
  • Sub-plan 03 ({phase}) — complete
  • All acceptance criteria met
  • Feature-wide manual validation passed
  • Ready for /commit

NOTES

Key Design Decisions

  • {Why decomposition over single plan}
  • {Why this phase breakdown}

Risks

  • {Risk 1 and mitigation}
  • {Risk 2 and mitigation}

Confidence Score: {X}/10

  • Strengths: {what’s clear and well-defined}
  • Uncertainties: {what might change or cause issues}
  • Mitigations: {how we’ll handle the uncertainties}

Graph View