en un clic
plan-milestone
// Break a milestone into discrete tasks. Use when starting work on a milestone that has been defined in the project plan.
// Break a milestone into discrete tasks. Use when starting work on a milestone that has been defined in the project plan.
Add a new AI coding agent to Agent Sandbox. Creates all required files (Dockerfile, templates, CI, docs) and wires the agent into the CLI, proxy, and build system.
Create a project plan for a new initiative. Use when starting a new project, defining goals, breaking work into milestones, and doing initial technical due diligence.
Entry point for the three-tier planning system. Use when starting planning work at any level - project, milestone, or task.
Plan and track execution of a task. Use when starting implementation of a task defined in a milestone plan.
| name | plan-milestone |
| description | Break a milestone into discrete tasks. Use when starting work on a milestone that has been defined in the project plan. |
Part of the three-tier planning system. See /plan for an overview.
This skill breaks a milestone into discrete, implementable tasks. The output is a milestone plan that sequences work and identifies dependencies.
Before starting, read:
docs/plan/project.md - to understand the milestone's context and goalsdocs/plan/learnings.md - to incorporate lessons from previous workdocs/plan/decisions/The final artifact is docs/plan/milestones/{milestone}/milestone.md.
The milestone plan can be updated as work progresses. When task scope changes significantly, update the task definition but also record the change in a Changes section. This preserves history for milestone retrospectives while keeping the plan current.
Work through these phases interactively. Each phase should be a conversation.
Review the milestone's definition from the project plan.
Confirm understanding before proceeding.
Review accumulated learnings that may apply to this milestone.
Note applicable learnings in the milestone plan.
Identify the discrete pieces of work.
A task should map to a well-crafted PR: a piece of functionality or a fix, along with relevant tests, sized appropriately for human peer review. Not so large that review becomes burdensome, not so small that it lacks meaningful context.
Understand how tasks relate to each other.
Flag uncertainties specific to this milestone.
Order the tasks for execution.
When all phases are complete, compile into docs/plan/milestones/{milestone}/milestone.md:
# Milestone: {identifier} - {name}
## Goal
{What this milestone achieves}
## Scope
{What's included and explicitly excluded}
## Applicable Learnings
{Lessons from previous work that apply here}
## Tasks
### {m1.1-task-name}
**Summary:** {One-sentence description of what this task delivers}
**Scope:**
- {What's included}
- {What's explicitly excluded or deferred}
**Acceptance Criteria:**
- {Condition that must be true when complete}
- {Another condition}
**Dependencies:** {What must be done first}
**Risks:** {Uncertainties or concerns}
### {m1.2-task-name}
**Summary:** {One-sentence description of what this task delivers}
**Scope:**
- {What's included}
- {What's explicitly excluded or deferred}
**Acceptance Criteria:**
- {Condition that must be true when complete}
- {Another condition}
**Dependencies:** {What must be done first}
**Risks:** {Uncertainties or concerns}
{Continue for all tasks}
## Execution Order
{Recommended sequence, noting parallelization opportunities and decision points}
## Risks
{Milestone-level risks and mitigation strategies}
## Definition of Done
{How we know this milestone is complete}
## Changes
{Record significant changes to scope or tasks as they occur}
### {Date}: {Change summary}
{What changed and why}