ワンクリックで
map
Analyze existing codebase for brownfield project support
Codex または Claude でインストール この Prompt をコピーして Codex、Claude、または他のアシスタントに貼り付けると、Skill ページを確認してインストールできます。
メニュー
Analyze existing codebase for brownfield project support
Codex または Claude でインストール この Prompt をコピーして Codex、Claude、または他のアシスタントに貼り付けると、Skill ページを確認してインストールできます。
SOC 職業分類に基づく
Initialize a Megazord project with configuration and planning structure
Systematic four-phase debugging (reproduce, isolate, root-cause, fix)
Socratic brainstorming to explore approaches before implementation
Execute the current phase plan (subagent or Agent Teams mode)
Show all available Megazord skills with descriptions and usage examples
Complete milestone lifecycle -- audit, archive, deferred items, and next version preparation
| name | map |
| description | Analyze existing codebase for brownfield project support |
| disable-model-invocation | false |
Analyze an existing codebase by spawning parallel mapper agents that produce structured analysis documents. Enables brownfield project onboarding by understanding architecture, tech stack, code quality, and concerns.
Reference @skills/init/design-system.md for visual output formatting.
Reference @skills/map/mapper.md for agent spawning patterns.
Output the stage banner:
╔═══════════════════════════════════════════════╗
║ ⚡ MEGAZORD ► MAP ║
╚═══════════════════════════════════════════════╝
Read .planning/megazord.config.json. If missing, display an error box and stop:
╔═══════════════════════════════════════════════╗
║ ✗ Project Not Initialized ║
╠═══════════════════════════════════════════════╣
║ No megazord.config.json found. ║
║ Run /mz:init to set up your project first. ║
╚═══════════════════════════════════════════════╝
If config exists, continue loading:
.planning/megazord.config.json for project settings.Determine the plugin path:
plugin_path from the config JSON.plugin_path is not set in config, try ~/.claude/plugins/mz. Check if ~/.claude/plugins/mz/bin/megazord.mjs exists.Plugin path not configured. Run
/mz:settingsand setplugin_path, or re-run/mz:init.
Parse the user's message for arguments:
/mz:map architecture, /mz:map tech, etc.arch -> architectureconventions -> qualitytech, architecture, quality, concernsDisplay the target:
▸ Target
{focus area name, or "All areas" if no focus specified}
Check if .planning/codebase/ directory exists and contains analysis documents.
Use Bash to check: ls .planning/codebase/*.md 2>/dev/null | wc -l
If existing documents found AND no focus parameter was given:
Use AskUserQuestion to offer 3 options:
Handle each choice:
.planning/codebase/ directory and recreate it: rm -rf .planning/codebase/If existing documents found AND focus parameter was given:
Use AskUserQuestion:
Handle each choice:
If no existing documents: Proceed to mapping.
mkdir -p .planning/codebase/
Read the mapper agent definition ONCE and reuse for all spawns:
Read {plugin_path}/agents/mz-mapper.md -> mapper_instructions
Before spawning, determine the model for mapper agents:
model_profile and model_overrides from the loaded config.model_overrides.mapper is set and not "inherit". If so, use that value. Otherwise, use the profile mapping: quality->opus, balanced->sonnet, budget->haiku.{plugin_path}/agents/mz-mapper.md frontmatter model field to the resolved value. Use simple string replacement to rewrite the model: {value} line.For each selected focus area, spawn one Task tool agent. Spawn ALL agents in parallel (or just the focused one if a focus parameter was given).
For each agent, the Task prompt must include:
<focus> tags<output_dir> tagsImportant: Use subagent_type="mz-mapper" for all agents. The agent definition is loaded from the registered agent file (frontmatter already updated with the resolved model). See @skills/map/mapper.md for the exact prompt structure.
Fallback: If spawning with subagent_type="mz-mapper" fails, fall back to subagent_type="general-purpose" with the agent definition embedded inline in <agent_role> tags.
Display progress for each active agent:
▸ Mapping
◆ Mapping tech...
◆ Mapping architecture...
◆ Mapping quality...
◆ Mapping concerns...
Wait for all agents to complete.
After completion, verify that expected output files exist. Use Bash to check file presence and line counts -- do NOT read the file content:
wc -l .planning/codebase/STACK.md .planning/codebase/INTEGRATIONS.md .planning/codebase/ARCHITECTURE.md .planning/codebase/STRUCTURE.md .planning/codebase/CONVENTIONS.md .planning/codebase/TESTING.md .planning/codebase/CONCERNS.md 2>/dev/null
Update the progress display with results:
▸ Mapping
✓ tech (STACK.md, INTEGRATIONS.md)
✓ architecture (ARCHITECTURE.md, STRUCTURE.md)
✓ quality (CONVENTIONS.md, TESTING.md)
✓ concerns (CONCERNS.md)
If any expected files are missing, display a warning but continue:
⚠ Missing: {filename} -- agent may have failed
Only run synthesis when ALL 4 areas were mapped (not on focused single-area runs).
Spawn the synthesis agent (model is already set from the frontmatter update in Step 5):
Task(
prompt="<focus>synthesis</focus>
<output_dir>.planning/codebase/</output_dir>
<instructions>Read all 7 documents in .planning/codebase/ (STACK.md, INTEGRATIONS.md,
ARCHITECTURE.md, STRUCTURE.md, CONVENTIONS.md, TESTING.md, CONCERNS.md).
Produce a compact executive SUMMARY.md with cross-cutting insights.
Write SUMMARY.md to the output directory. Return confirmation only.</instructions>",
subagent_type="mz-mapper",
description="Map codebase: synthesis"
)
Fallback: If spawning with subagent_type="mz-mapper" fails, fall back to subagent_type="general-purpose" with the agent definition embedded inline in <agent_role> tags.
Display:
▸ Synthesis
◆ Synthesizing findings...
Wait for completion, then verify SUMMARY.md exists:
wc -l .planning/codebase/SUMMARY.md 2>/dev/null
Update display:
▸ Synthesis
✓ SUMMARY.md produced
Display the results action box showing all produced documents:
╔═══════════════════════════════════════════════╗
║ Codebase Mapped ║
╠═══════════════════════════════════════════════╣
║ ║
║ ▸ Documents ║
║ STACK.md ({N} lines) ║
║ INTEGRATIONS.md ({N} lines) ║
║ ARCHITECTURE.md ({N} lines) ║
║ STRUCTURE.md ({N} lines) ║
║ CONVENTIONS.md ({N} lines) ║
║ TESTING.md ({N} lines) ║
║ CONCERNS.md ({N} lines) ║
║ SUMMARY.md ({N} lines) ║
║ ║
║ ▸ Areas Mapped ║
║ ✓ tech ✓ architecture ║
║ ✓ quality ✓ concerns ║
║ ║
╚═══════════════════════════════════════════════╝
For focused runs, only show the documents from the mapped area.
End with the Next Up block:
═══════════════════════════════════════════════════
▸ Next Up
**Plan from analysis** -- create roadmap from codebase map
`/mz:plan`
═══════════════════════════════════════════════════
/mz:init. Exit./mz:settings. Exit.wc -l).{plugin_path} resolution from config (same pattern as /mz:plan Step 2).arch for architecture, conventions for quality.