skills.rest
tapway
tapwayOFFICIAL
Explore all skills by this author →
WebsiteGitHub

plan

Writes a timestamped markdown implementation plan to .hermes/plans/ without executing code.

1|Updated Jun 21, 2026
Search Tags:#workflow#planning#implementation plan#approval gate#plan mode#hermes#markdown plan
One-click install
npx skills add https://github.com/tapway/shogun-os --skill plan-tapway
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: plan
Source: https://github.com/tapway/shogun-os/tree/main/skills/hermes/plan
Command: npx skills add https://github.com/tapway/shogun-os --skill plan-tapway

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? It prevents premature implementation by enforcing a structured planning phase between brainstorming and execution, producing a concrete, reviewable markdown plan before any code changes happen. ## Core Features & Use Cases - Planning-only mode: Inspects the repository with read-only commands and writes a plan without editing project files, running mutating commands, or committing changes. - Mandatory three-phase gate: Enforces Brainstorm → Plan → Execute, blocking planning if brainstorming never happened and blocking execution until the user approves the plan. - Timestamped plan output: Saves plans as .hermes/plans/YYYY-MM-DD_HHMMSS-<slug>.md relative to the active workspace, working across local, docker, ssh, modal, and daytona backends. - Use Case: Before adding user authentication, invoke /plan add user authentication to get a saved plan covering the goal, approach, files to change (e.g., src/auth.py, tests/test_auth.py), risks, and verification steps, then approve it before any code is written. ## Quick Start Ask the agent to plan a task with a command like "/plan add user authentication" and it will inspect the repo read-only, save a detailed markdown plan under .hermes/plans/, and wait for your approval before executing.

Frequently Asked Questions about plan

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

FAQPage Schema
How do I make an AI agent write a plan before coding?▼

Invoke the /plan slash command with your task, such as "/plan add user authentication". The agent inspects the repository with read-only commands, writes a concrete markdown plan to .hermes/plans/, and waits for your explicit approval before executing anything.

Where are plan files saved by the plan skill?▼

Plans are saved under .hermes/plans/ in the active workspace using the filename format YYYY-MM-DD_HHMMSS-<slug>.md. The relative path works across local, docker, ssh, modal, and daytona backends.

Can the plan skill modify code or run commands?▼

No. In plan mode the agent cannot implement code, edit project files other than the plan markdown, run mutating terminal commands, commit, push, or perform external actions. It may only use read-only inspection tools.

What happens if I skip brainstorming and go straight to planning?▼

The skill enforces a hard gate: if no brainstorming phase occurred for the task, it stops and refuses to write a plan. It flags that brainstorming must come first and offers to start that phase.

When should I not use plan mode?▼

Skip plan mode for simple one-file fixes that can be done directly, tasks that have not been brainstormed yet, and the execution phase itself, where you should follow the approved plan rather than re-planning.

Related Skills

Explore More Software Engineering Skills→
lucasFR2lucasFR2

brainstorming

Guide structured brainstorming sessions with clarifying questions and design proposals.

Community
Advanced
atomicinnovationatomicinnovation
30

implement-plan

Execute approved implementation plans and update progress checkboxes.

Official
Advanced
hellowind777hellowind777
702

hello-subagent

Orchestrates subagent task delegation with context isolation and two-stage review.

Community
Intermediate
FerroxLabsFerroxLabs
21

ferrox-fast

Apply trivial low-risk code changes directly without planning.

Community
Basic