skills.rest
speakeasy-api
speakeasy-apiOFFICIAL
Explore all skills by this author →
Website𝕏@speakeasydevGitHub

granary-execute-task

Execute Granary sub-agent tasks by building context and reporting progress.

19|2|Updated Jan 12, 2026
Search Tags:#orchestration#handoff#task-management#sub-agent#execute-task#granary
One-click install
npx skills add https://github.com/speakeasy-api/granary --skill granary-execute-task
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: granary-execute-task
Source: https://github.com/speakeasy-api/granary/tree/main/.claude-plugin/skills/execute-task
Command: npx skills add https://github.com/speakeasy-api/granary --skill granary-execute-task

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill enables a sub-agent to execute a designated Granary task by building the necessary task context, applying the required changes, and systematically reporting progress back to the orchestrator.

Core Features & Use Cases

  • Context-driven execution: The sub-agent builds task context using granary handoff to understand goals, files to modify, implementation details, and steering conventions.
  • Automated task execution: It reads the handoff outputs, implements changes, and verifies completion according to acceptance criteria.
  • Progress reporting & blockers: The sub-agent records progress with granary task comments and signals blockers or completion to the orchestrator.

Quick Start

Start a new task by obtaining its handoff and then proceed through the lifecycle:

  1. granary handoff --tasks <task-id>
  2. granary task <task-id> start
  3. Implement the required changes as described in the handoff and periodically update progress with granary task <task-id> comments create "Started work"
  4. granary task <task-id> done --comment "Completed task per handoff"

Frequently Asked Questions about granary-execute-task

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I execute a specific task in a multi-agent orchestration workflow?▼

To execute a specific task in a multi-agent orchestration workflow, a sub-agent reads the handoff context using a task-id, implements the required file modifications, and signals completion or blockers back to the orchestrator.

What context does a sub-agent need to start an orchestrated task?▼

A sub-agent needs task details, files to modify, implementation details, and steering files obtained from the granary handoff process to build the necessary context before starting an orchestrated task.

How do I report progress and signal blockers during task execution?▼

You report progress and signal blockers during task execution by recording periodic updates with task comments and using the task done command to signal either completion or any blockers to the orchestrator.

Does multi-agent task execution support safety features like leases?▼

Yes, multi-agent task execution supports safety features like leases and progress updates to ensure coordinated safety and prevent conflicts when multiple sub-agents operate within the same orchestrated workflow.

What is the best way to manage handoffs between an orchestrator and a sub-agent?▼

The best way to manage handoffs between an orchestrator and a sub-agent is using a structured handoff command to pass task details and steering files, ensuring the sub-agent understands the goals before implementing changes.

Related Skills

Explore More Software Engineering Skills→
speakeasy-apispeakeasy-api
19

granary-setup

Install Granary, initialize the workspace, and verify readiness with granary doctor.

Official
Intermediate
duongdevduongdev
9

ccpm-code-review

Verify tests, builds, checklists, and blockers before CCPM task completion or PR creation.

Community
Intermediate
speakeasy-apispeakeasy-api
19

granary-orchestrate

Coordinate sub-agents to execute parallel tasks with granary commands.

Official
Advanced
ryan-nguyen-01ryan-nguyen-01
4

agent-orchestrator

Coordinate multi-agent workflows by analyzing tasks and spawning context-aware agents.

Community
Advanced