plan
SkillDev toolsplan is a skill that lets your agent break a task into a step-by-step plan before doing it. It supports lightweight planning with an optional interview, so the agent can ask clarifying questions first when needed. Use it when you want the agent to think through the work before making changes.
Use plan in Claude, ChatGPT or Ahel Desktop
Free. Sign in, add plan and connect your AI. About a minute.
Also: Claude Code · Cursor · Codex
Then ask your AI: use the plan skill
Details
Instructions available. Your AI can read the instructions. Execution depends on the setup they require.
Account requirements not reviewed. Check the skill instructions before use; ahel provides instructions and does not run this skill.
No other account needed.
Make sure your agent supports skills.
What your AI can do with it
- Break a task into a step-by-step plan
- Run an optional interview to gather details
- Keep planning lightweight and focused
- Produce a plan before the agent starts work
Getting started
- Make sure your agent supports skills.
- Add the plan skill to your agent's configuration.
- Give your agent a task and ask it to use plan.
- Answer any interview questions if the agent asks them.
- Review the step-by-step plan before the agent proceeds.
What this skill tells your AI
The instructions your AI receives, as published by yeachan-heo/oh-my-codex in skills/plan/SKILL.md and read by ahel’s review.
<Use_When>
- Planning before implementation — "plan this", "let's plan"
- Requirements are vague and need scoping — use
--interview - Reviewing an existing plan —
--review</Use_When>
<Do_Not_Use_When>
- Want direct execution — execute the scoped task directly; use
ultragoalonly for durable goals orteamfor coordinated parallel work - Simple focused fix with obvious scope — skip planning </Do_Not_Use_When>
<Execution_Policy>
- Auto-detect interview vs direct;
--interviewforces Socratic mode,--directskips it - Ask one question at a time; gather codebase facts first
omx exploreis deprecated. Use normal repository inspection for codebase facts;omx sparkshellonly for explicit shell-native evidence- Plans must be evidence-backed: 80%+ claims cite file/line, 90%+ criteria testable
- Right-size steps to scope; save to
.omx/plans/</Execution_Policy>
Mode Selection
| Mode | Trigger | Behavior |
|---|---|---|
| Interview | --interview or broad request | Socratic requirements gathering |
| Direct | --direct or detailed request | Generate plan immediately |
| Review | --review | Critic evaluation of existing plan |
Interview (--interview)
- Classify request; ask one focused question via
omx question - Gather codebase facts first, then ask informed follow-ups
- Build on answers; consult Analyst for hidden requirements
- Create plan when user signals readiness
Direct
- Quick analysis (optional Analyst)
- Generate plan immediately
- Optional Critic review if requested
Review (--review)
- Read plan from
.omx/plans/ - Evaluate via Critic; verdict APPROVED/REVISE/REJECT
- If author == reviewer, hand off to
$code-review
Output
Every plan includes: Requirements Summary, Acceptance Criteria (testable), Implementation Steps (with file refs, adaptive count), Risks/Mitigations, Verification. Saved to .omx/plans/.
Outcome-first framing: apply outcome-first framing, concise visible updates for multi-step planning. Local overrides for the active workflow branch apply via newer user task updates. If the user says continue, continue the current branch instead of restarting.
<Tool_Usage>
omx exploreis deprecated. Use normal repository inspection;omx sparkshellonly for explicit shell-native evidence- Use
omx questionfor interview questions; plain text only as fallback - Use
ask_codexwithplanner/analyst/criticas needed </Tool_Usage>
<Escalation_And_Stop_Conditions>
- Stop interviewing when requirements are clear — do not over-interview
- Escalate on irreconcilable trade-offs requiring business decision
- "just do it" / "skip planning" → hand off to
$ultragoal</Escalation_And_Stop_Conditions>
<Final_Checklist>
- Acceptance criteria testable (90%+ concrete)
- Claims cite file/line (80%+)
- Risks have mitigations, no vague terms
- Plan saved to
.omx/plans/</Final_Checklist>
Review Quality Criteria
| Criterion | Standard |
|---|---|
| Clarity | 80%+ claims cite file/line |
| Testability | 90%+ criteria concrete |
| Verification | File refs exist |
Signals
- GitHub stars
- 33k
- Forks
- 3k
- Last commit
- Oct 2026
Questions
- What is plan?
- plan is a skill that lets your agent break a task into a step-by-step plan before doing it. It supports lightweight planning with an optional interview.
- Does plan always ask interview questions?
- No. The interview is optional. The agent can plan without asking questions, or it can use the interview to gather more details when needed.
- When should I use plan?
- Use plan when you want your agent to think through a task and produce a step-by-step plan before making changes.
- Can plan run without an interview?
- Yes. The interview is optional, so the agent can create a plan directly without asking clarifying questions.
Advanced
- Item type
- skill
- Key
plan-yeachan-heo- Source
- github.com/yeachan-heo/oh-my-codex
github.com/yeachan-heo/oh-my-codex