run-blueprint

Mode d’emploi

Copiez la commande et examinez le paquet avant de l’installer dans votre environnement agent.
Consultez le document original pour les outils, identifiants et environnements requis.
Décrivez les entrées et le livrable souhaité ; comparez la sortie aux sources.

npx skills add https://github.com/majesticlabs-dev/majestic-marketplace --skill run-blueprint

Avant de commencer

L’installation n’inclut ni abonnement au modèle, ni quota API, ni accès aux fichiers.
Étoiles et installations reflètent l’activité, pas un test indépendant de qualité ou de sécurité.

Document Skill original

Rendu depuis le dépôt source en conservant titres, exemples, code, tableaux, liens et images.

Run Blueprint

Execute all tasks in a blueprint plan sequentially, respecting dependencies.

Input

plan_path: $ARGUMENTS (required)

Task Tracking Setup

TASK_TRACKING = config_read("task_tracking.enabled", "false")

If TASK_TRACKING:
  BLUEPRINT_ID = "run-blueprint-{timestamp}"
  TASK_MAP = {}  # Maps T1, T2 etc to Claude Task IDs

Workflow

1. Load Blueprint

Read(file_path: plan_path)

Parse the ## Implementation Tasks section:

ID	Task	Points	Dependencies	Status
T1	Setup auth middleware	2	-	⏳
T2	Create login endpoint	2	T1	⏳

Task status icons:

⏳ Pending
🔄 In Progress
✅ Completed
🔴 Failed

1.5. Create Tasks from Table (if tracking enabled)

If TASK_TRACKING:
  For each ROW in implementation_tasks:
    TASK = TaskCreate:
      subject: "{ROW.id}: {ROW.task}"
      activeForm: "Building {ROW.task}"
      metadata: {blueprint_id: BLUEPRINT_ID, task_id: ROW.id, points: ROW.points, source: plan_path}
    TASK_MAP[ROW.id] = TASK.id

  # Set up dependencies
  For each ROW in implementation_tasks where ROW.dependencies != "-":
    BLOCKED_BY = [TASK_MAP[dep] for dep in ROW.dependencies.split(",")]
    TaskUpdate(TASK_MAP[ROW.id], addBlockedBy: BLOCKED_BY)

2. Find Next Task

If TASK_TRACKING:
  TASKS = TaskList()
  BLUEPRINT_TASKS = filter(TASKS, task.metadata.blueprint_id == BLUEPRINT_ID)
  NEXT_TASK = first(BLUEPRINT_TASKS where status == "pending" AND blockedBy.length == 0)
Else:
  For each task in Implementation Tasks:
    If task.status == ⏳:
      If task.dependencies all ✅:
        NEXT_TASK = task
        Break

If no task found (all ✅ or 🔴): Go to Step 5.

3. Execute Task

Skill("build-task") "{task.id}" --no-ship

4. Update Status

If TASK_TRACKING:
  CLAUDE_TASK_ID = TASK_MAP[NEXT_TASK.id]
  TaskUpdate(CLAUDE_TASK_ID, status: "in_progress")
  # After build completes
  If build succeeded: TaskUpdate(CLAUDE_TASK_ID, status: "completed")
  If build failed: TaskUpdate(CLAUDE_TASK_ID, status: "completed", metadata: {result: "failed"})

# Always update markdown (for visibility in file)
Edit(file_path: plan_path):
  If build succeeded: task.status = ✅
  If build failed: task.status = 🔴

Go to Step 2 (next task).

5. Ship

When all tasks are ✅ or 🔴:

Skill("ship-it")

Task ID Formats

System	Format	Example
GitHub Issues	`#{number}`	`#123`
Linear	`{PROJECT}-{number}`	`PROJ-123`
Beads	`BEADS-{number}`	`BEADS-123`
File-based	`TODO-{number}`	`TODO-123`

Error Handling

Scenario	Action
Task build fails	Mark 🔴, continue to next task
All dependencies failed	Skip task, mark 🔴
Blueprint file not found	Error, exit
No tasks section found	Error, exit