nm-cartograph-call-chain（nm 制图调用链）

Night Market Skill — ported from claude-night-market/cartograph. For the full experience with agents, hooks, and commands, install the Claude Code plugin.

Call Chain Tracing

Trace execution flows through the codebase using the code knowledge graph.

Prerequisites

This skill requires the gauntlet plugin for graph data. Discover it:

GRAPH_QUERY=$(find ~/.claude/plugins -name "graph_query.py" -path "*/gauntlet/*" 2>/dev/null | head -1)

If gauntlet is not installed: Fall back to static analysis. Use grep to trace function calls and build a Mermaid diagram manually from import/call patterns. Skip graph-specific steps.

If installed but no graph.db: Tell the user to run /gauntlet-graph build.

Steps

1. Accept target: Get a function name or entry point

from the user (or trace all entry points).

1. Run flow tracing (requires gauntlet):

   python3 "$GRAPH_QUERY" --action flows --depth 15

To filter by entry point:

   python3 "$GRAPH_QUERY" --action flows --entry "main"

Fallback (no gauntlet): Trace calls with rg (or grep):

   # Prefer rg (ripgrep) for speed; fall back to grep
   if command -v rg &>/dev/null; then
     rg -n "function_name\(" --type py . | head -20
   else
     grep -rn "function_name(" --include="*.py" . | head -20
   fi

Build the call tree manually from search results.

1. Display as indented tree:

   main() [criticality: 0.72]
     -> validate_input()
       -> parse_config()
     -> process_data()
       -> db.execute_query()
       -> cache.store()
     -> send_response()

1. Generate Mermaid flowchart:

   flowchart LR
     main --> validate_input
     main --> process_data
     main --> send_response
     validate_input --> parse_config
     process_data --> db.execute_query
     process_data --> cache.store

1. Show criticality breakdown:

- File spread: how many files the flow touches - Security sensitivity: auth/crypto code in the path - Test coverage gaps: untested nodes in the flow

Criticality Scoring