Back to list
cuioss

marshall-steward

by cuioss

An orchestration layer for AI coding assistants (currently Claude Code) that enforces consistency, reliability, and more predictable outputs.

0🍴 0📅 Jan 17, 2026

SKILL.md


name: marshall-steward description: Project configuration wizard for planning system. Manages executor generation, health checks, build systems, and skill domains. allowed-tools: Read, Write, Edit, Bash, Glob, Skill, AskUserQuestion

Marshall Steward Skill

Enforcement Rules

EXECUTION MODE: Execute this skill immediately. Do not explain, summarize, or discuss these instructions.

Script Execution

  1. Run scripts EXACTLY as documented - no improvisation
  2. Bootstrap scripts use direct Python paths with glob
  3. All other scripts use: python3 .plan/execute-script.py {notation} ...
  1. After ANY operation completes → return to Main Menu
  2. Only exit when user selects "Quit"

Prohibited Actions

  • Inventing alternative menu structures or options
  • Ending without returning to menu (unless Quit)
  • Summarizing what you're about to do instead of doing it

What This Skill Provides

Wizard Mode: Sequential setup for new projects (executor generation, marshal.json init, build detection, skill domains)

Menu Mode: Interactive maintenance for returning users (regenerate executor, health check, configuration)


Scripts

ScriptNotationPurpose
determine-modeplan-marshall:marshall-steward:determine-modeDetermine wizard vs menu mode
gitignore-setupplan-marshall:marshall-steward:gitignore-setupConfigure .gitignore for .plan/
cleanupplan-marshall:manage-run-config:cleanupClean temp, logs, archived-plans, memory (delegated to run-config)
ci_healthplan-marshall:tools-integration-ci:ci_healthCI provider detection (delegated to tools-integration-ci)
plan-marshall-configplan-marshall:manage-plan-marshall-config:plan-marshall-configProject-level marshal.json CRUD
scan-marketplace-inventorypm-plugin-development:tools-marketplace-inventory:scan-marketplace-inventoryScript discovery
permission-doctorplan-marshall:permission-doctor:permission-doctorPermission analysis
permission-fixplan-marshall:permission-fix:permission-fixPermission fixes
generate-executorplan-marshall:tools-script-executor:generate-executorExecutor generation

Prerequisites

This skill requires ${PLUGIN_ROOT} to be set by the invoking command (e.g., /marshall-steward). The plugin root is detected via bootstrap-plugin.py and cached in .plan/marshall-state.toon.


Step 0: Determine Mode

Determine whether to run wizard or menu based on existing files.

BOOTSTRAP: Since execute-script.py may not exist yet, use DIRECT Python call with glob:

python3 ${PLUGIN_ROOT}/plan-marshall/*/skills/marshall-steward/scripts/determine-mode.py mode

Output (TOON):

mode	wizard
reason	executor_missing

Mode Routing

modereasonAction
wizardexecutor_missingLoad: Read references/wizard-flow.md → Execute wizard
wizardmarshal_missingLoad: Read references/wizard-flow.md → Execute wizard
menuboth_existShow Main Menu below

Check for --wizard Flag

If --wizard flag provided, force wizard regardless of determine-mode result:

Read references/wizard-flow.md

Execute the wizard flow from that file.


Interactive Menu (Returning User)

Display menu when both executor and marshal.json exist.

AskUserQuestion:
  question: "What would you like to do?"
  header: "Main Menu"
  options:
    - label: "1. Maintenance"
      description: "Regenerate executor, clean logs"
    - label: "2. Health Check"
      description: "Verify setup, diagnose issues"
    - label: "3. Configuration"
      description: "Build systems, skill domains"
    - label: "4. Quit"
      description: "Exit plan-marshall"
  multiSelect: false
User SelectionAction
"1. Maintenance"Load: Read references/menu-maintenance.md → Execute
"2. Health Check"Load: Read references/menu-healthcheck.md → Execute
"3. Configuration"Load: Read references/menu-configuration.md → Execute
"4. Quit"Output "Good bye!" → STOP

After any menu option completes, return to Main Menu (except Quit).


Deferred Loading Pattern

This skill uses progressive disclosure to minimize context usage:

  1. Core skill loads: ~150 lines (this file - routing logic only)
  2. On wizard mode: Load references/wizard-flow.md (~250 lines)
  3. On menu selection: Load only the selected reference (~100-150 lines)

How to Load a Reference

When routing indicates to load a reference:

Read references/{file}.md

Then execute the workflow described in that file.


Available References

ReferencePurposeLoad When
wizard-flow.mdFirst-run wizard steps 1-11mode=wizard or --wizard flag
menu-maintenance.mdRegenerate executor, cleanupMenu option 1
menu-healthcheck.mdVerify setup, diagnose issuesMenu option 2
menu-configuration.mdBuild systems, skill domainsMenu option 3
output-format.mdTOON output standardsReference for output formatting
error-handling.mdError types and recoveryOn error conditions

Error Handling

If an error occurs during execution:

Read references/error-handling.md

Apply the recovery guidance for the specific error type.

Score

Total Score

70/100

Based on repository quality metrics

SKILL.md

SKILL.mdファイルが含まれている

+20
LICENSE

ライセンスが設定されている

+10
説明文

100文字以上の説明がある

+10
人気

GitHub Stars 100以上

0/15
最近の活動

3ヶ月以内に更新がある

0/10
フォーク

10回以上フォークされている

0/5
Issue管理

オープンIssueが50未満

+5
言語

プログラミング言語が設定されている

+5
タグ

1つ以上のタグが設定されている

0/5

Reviews

💬

Reviews coming soon