| name | sprang-explain |
| description | Deep-dive explanation of a specific file, function, or module. Use when the user says "/sprang-explain", "explain this file", "what does this function do", or "deep dive on X". |
Deep-dive explanation of a specific file, function, or module.
Arguments: <file path or path:functionName>
- Call
sprang_query with the argument to find the target node.
- Call
sprang_node with the resolved node_id to get: layer (name + id), layer_mate_count, in_degree, out_degree, has_annotation, and full 1-hop neighborhood.
- Call
sprang_why to retrieve decision context: commit history, primary authors, rationale snippets, change frequency, and any team annotation.
- Read the actual source file for current state.
- Produce a structured explanation:
- What it does — purpose and responsibilities (2-3 paragraphs)
- Where it fits — layer name, imports (outgoing), used by (incoming)
- How it works — key functions/classes with line number references
- History — who wrote it, when, why (from decision_context)
- Risk — risk_score, blast radius, structural warnings
- Next steps — suggest edits only after checking
risk_score
$ARGUMENTS
