Explain complex code

Explain Complex Code

Relies on BDK foundation (STARTUP_INSTRUCTIONS.md) for project context and MCP tool preference.

Generate architecture docs for complex code with visual diagrams and examples.

Workflow

1. Understand Scope

path arg specifies code to explain. Clarify:

• Focus areas: Architecture, algorithms, data flow, testing strategy
• Audience level: Team onboarding, external docs, debugging guide

2. Analyze Code Structure

Step 2.1: Initial Discovery

!python3 ${CLAUDE_PLUGIN_ROOT}/scripts/inject.py --chain ${CLAUDE_PLUGIN_ROOT}/fragments/tool-tiers/explore.chain.json

Using the exploration tools above: find all symbols, identify high-dependency symbols, use community grouping to determine subagent partitioning where available.

Step 2.2: Map Dependencies For each key class/function: query_graph(pattern="callers_of", node=<symbol>) and query_graph(pattern="callees_of", node=<symbol>). Fall back to Serena find_referencing_symbols if graph unavailable.

Step 2.3: Decide Partitioning Strategy

When graph communities are available, use community boundaries as partitioning units — one community per subagent. Otherwise fall back to file-count rules:

Files Count	Strategy	Subagent Count
1-2 files	One subagent for entire module	1 subagent
3-5 files	Group into 1-2 logical blocks	1-2 subagents
6-10 files	Group into 2-3 logical blocks	2-3 subagents
10+ files	Group by architectural layers	3-4 subagents

Never more than 3-4 subagents.

3. Launch Subagents Strategically

Launch ALL subagents in ONE message. Per subagent:

Explore the following files as a logical block: [FILE_LIST]

For EACH file, report:
1. Core responsibilities and purpose
2. Key classes/functions and their roles
3. Important dependencies
4. Critical algorithms or rules
5. Data structures
6. Edge cases or special handling

Then synthesize:
7. How these files work together
8. Data flow between them
9. Shared dependencies or patterns

4. Synthesize Documentation

From subagent findings, create structured docs.

Structure:

1. Overview: 2-3 sentence summary
2. Core Architecture: File tree + component diagram
3. Architecture Flow: Process flow with Graphviz
4. Critical Rules: Non-obvious rules with examples
5. Live Examples: Prototype code (NOT actual implementation)
6. Core Classes: Key classes with usage patterns
7. Testing Coverage: Existing test structure and patterns

5. Create Graphviz Diagrams

Use for:

• Data flow: How data moves through system
• Component architecture: Layers and dependencies
• Algorithm flow: Decision points and steps

Under 15 nodes. Label edges clearly.

6. Write Prototype Examples

CRITICAL RULE: Prototype code only — NOT actual implementation.

Good Prototype:

# Parse and validate
parsed = parse(input_data)
if not parsed.is_valid:
    return error_response()

# Transform based on type
result = transform(parsed)
return format_output(result)

Prototype Rules:

• Placeholder function names
• Clear control flow
• Comments for key steps
• Under 20 lines per example
• One concept per example

7. Save Documentation

Save to .bdk/explain-complex-code/[feature-name].md.

8. Compile Diagrams

After saving, invoke /bdk:graphviz-docs-compiler to extract .dot files and compile to SVG.

Quality Checklist

• Overview (2-3 sentences)
• File tree
• At least one Graphviz diagram
• Dot code blocks replaced with SVG image references
• Critical rules with examples
• Live examples with prototype code
• Core classes with usage
• Testing coverage

Notes

• Prototype code mandatory: Never copy actual implementation
• Parallel subagents save time: Launch all in one message
• Focus non-obvious details: Skip basic patterns