tiann/execplan-skill
An [Agent Skill](https://agentskills.io) that enables AI coding agents to tackle complex, long-running implementation tasks autonomously.
ExecPlan is an Agent Skill designed to help AI coding agents autonomously manage complex, long-running implementation tasks. It utilizes a structured execution plan that allows users to review and approve the approach before implementation starts. The system tracks progress in real-time and provides rollback procedures for safe recovery. It is compatible with tools such as Claude Code, Cursor, and Goose.
- Autonomous execution for long-running tasks
- Real-time tracking of progress and decisions
- Built-in rollback procedures for safe recovery
full readme from github
ExecPlan Skills
References:
- Codex Execution Plans — OpenAI Cookbook
- Video Introduction — YouTube
An Agent Skill that enables AI coding agents to tackle complex, long-running implementation tasks autonomously.
What You Get
- Extended autonomous execution — Agent works independently for hours without losing context
- Verifiable approach — Review and approve the plan before implementation begins
- Living documentation — Progress, decisions, and discoveries tracked in real-time
- Observable outcomes — Every milestone produces working, testable results
- Safe recovery — Built-in rollback procedures if something goes wrong
Installation
Add this skill to your agent's skills directory:
.skills/
└── execplan/
Compatibility
Works with: Claude Code, Cursor, OpenCode, Amp, Goose, VS Code, and other Agent Skills-compatible tools.
License
MIT