plan-mode
Use when a task needs a written execution plan with explicit steps, risks, verification, and user approval before implementation starts.
Skill by nexttylabs
Chalk Labs
Need help deploying plan-mode?
Chalk Labs sets up agents like this inside real teams: connected to your tools, tested on your data, with a human in the loop where it matters.
Install
Works with Claude Code, Claude apps, Codex, Cursor, Gemini CLI, GitHub Copilot, VS Code, OpenCode, Goose, Windsurf. Read the repository before you give any tool access to your data.
npx agentmarket install skill/nexttylabs-plan-mode
Asks which of your tools to add it to (Claude Code, Codex, Cursor, VS Code, Windsurf, Gemini CLI, Claude Desktop) and shows the change first. See https://agentmart.dev/cli.
Get the weekly diff
The weekly diff: one email a week with new agent tools, notable updates and grade changes, plus security alerts for tools you follow. Double opt-in, one-click unsubscribe.
About
Use cases: productivity
Reviews
Write a reviewNo reviews yet. Installed it? Reviews come from verified installers only. How reviews work
Permissions
What this skill can do and what it asks for, from its SKILL.md and code. Checked Nov 1, 2026.
- Tools
- No tools found
- Secrets
- None asked for
- Install scripts
- None found
Network hosts it talks to 1
codeload.github.com· found in install command
Files it touches
~/.claude/skills/plan-mode
What's inside
- Skill name
plan-mode- Description
- Use when a task needs a written execution plan with explicit steps, risks, verification, and user approval before implementation starts.
- License
- MIT
- Instructions
- 121 lines in SKILL.md · view on GitHub
Files (1)
- SKILL.md
SKILL.md (first lines)
# Plan Mode A structured lifecycle for breaking complex work into verifiable steps, writing them to a plan document, and executing them autonomously. ## 1. Creating a Plan ### Gather context Before writing anything, collect the information needed to plan well: - Read relevant files, configs, and documentation. - Identify constraints (time, risk, dependencies, available tools). - Clarify ambiguous requirements with the user. ### Analyze - Determine the minimal set of changes required. - Identify independent vs. sequential steps. - Flag anything that needs user approval before proceeding. ### Write the plan Create a file at `plans/<slug>.md` using the format below. The slug should be a short, lowercase, hyphenated description (e.g., `add-retry-logic`). ### Emit checklist After writing the plan file, print a numbered checklist summary so the user can review and approve it. ## 2. Plan Document Format ```markdown # Plan: <title> plan_id: <slug> status: draft | approved | in_progress | blocked | completed | abandoned current_focus: <step number or none> ## Goal One or two sentences describing the desired outcome. ## Success Criteria - [ ] Criterion 1 - [ ] Criterion 2 ## Steps ### Step 1 — <short description> - **Tools**: <which tools or commands this step uses> - **Risk**: low | medium | high - **Estimate**: <time or effort estimate> - **Details**: What to do and what "done" looks like. ### Step 2 — ... ## Risks - Risk description → mitigation strategy. ## Progress Log | Step | Status | Notes | |------|-------------|-------| | 1 | pending | | | 2 | pending | | ``` ## 3. Plan Rules 1. Each step must specify the **tools** it will use. 2. Each step must be **independently verifiable** — there should be a concrete way to confirm it succeeded. 3. Each step must include a **risk assessment** (low / medium / high). 4. Each step must include a **time estimate**. 5. A plan may have at most **20 steps**. If more are needed, split into multiple plans. 6. Steps should be ordered so that failures are caught early (high-risk or validation steps first when possible). ## 4. Approving and Executing ### Approve The plan stays in `draft` status until the user explicitly approves it. Do not begin execution on a draft plan. ### Execute 1. Read the plan file to determine the current state. 2. Set plan status to `in_progress`. 3. Update `current_focus` to the step you are actively working on. 4. For each step marked `pending`: - Check `current_focus` — only work on one step at a time. - Execute the step using the specified tools. - Verify the result. - Update the Progress Log: set status to `done`, `failed`, or `skipped`, and add notes. 5. When all steps are resolved, set the plan status to `completed` and `current_focus` to `none`. ### Handle failures - If a step fails, update its status to `failed` with a description of what went wrong. - Attempt the step once more if the failure seems transient. - If it fails again, set the plan status to `blocked` and report to the user. Do not skip ahead to later steps that depend on the failed one. ## 5. Checking Status Re-read the plan file and print the Progress Log table. Highlight any steps that are `failed` or `blocked`. ## 6. Listing Plans List all files in the `plans/` directory and print their `plan_id` and `status`. ## 7. Revising Plans If requirements change mid-execution: 1. Set the plan status to `draft`. 2. Update the Goal, Steps, and Risks as needed. 3. Reset any not-yet-started steps to `pending`. 4. Set `current_focus` to `none`. 5. Get user approval again before resuming execution.
From nexttylabs/kubebar, MIT.
More from nexttylabs/kubebar
- 0.598score
fix-issue
nexttylabs
Use when a GitHub issue should be handled end-to-end, including issue analysis, code changes, tests, verification, and an optional commit.
SkillAFreshness: A★ 2Claude Code +9 - 0.598score
github
nexttylabs
Use when working with GitHub repositories, issues, pull requests, branches, or commits and the task needs authenticated repository operations or GitHub…
SkillAFreshness: A★ 2Claude Code +9 - 0.598score
pr-shepherd
nexttylabs
Use when a pull request needs one coordinated workflow covering review state, fixes, checks, branch updates, and merge readiness.
SkillAFreshness: A★ 2Claude Code +9 - 0.598score
respond-pr
nexttylabs
Use when a pull request already has review feedback that needs triage, code changes, verification, and clear replies back to reviewers.
SkillAFreshness: A★ 2Claude Code +9 - 0.598score
review-checklist
nexttylabs
Use when deciding whether a change is ready to merge and a concise cross-language checklist is more useful than a full narrative review.
SkillAFreshness: A★ 2Claude Code +9 - 0.598score
review-pr
nexttylabs
Use when a pull request needs a serious pre-merge review covering correctness, edge cases, security, tests, documentation, and architecture.
SkillAFreshness: A★ 2Claude Code +9 - 0.598score
triage-issues
nexttylabs
Use when a GitHub issue backlog needs read-only classification, prioritization, duplicate spotting, or clarification gaps called out.
SkillAFreshness: A★ 2Claude Code +9 - 0.598score
triage-prs
nexttylabs
Use when open pull requests need a read-only dashboard showing review state, scope, blockers, conflicts, and merge readiness.
SkillAFreshness: A★ 2Claude Code +9 - 0.576score
ship
nexttylabs
Run the Swift iOS quality gate before shipping.
SkillAFreshness: A★ 2Claude Code +9
Signals
- Stars
- 2
- Forks
- 0
- Open issues
- 0
- Last commit (this folder)
- Apr 18, 2026 (6 months ago)
- Last release
- Oct 4, 2026 (28 days ago)
From GitHub (fetched Oct 24, 2026).
Details
- License
- MIT
- Repository
- github.com/nexttylabs/kubebar
- First seen
- Oct 24, 2026
- Last checked
- Oct 24, 2026
- Source
- GitHubgithub-graphql
More from nexttylabs
- 0.598score
fix-issue
nexttylabs
Use when a GitHub issue should be handled end-to-end, including issue analysis, code changes, tests, verification, and an optional commit.
SkillAFreshness: A★ 2Claude Code +9 - 0.598score
github
nexttylabs
Use when working with GitHub repositories, issues, pull requests, branches, or commits and the task needs authenticated repository operations or GitHub…
SkillAFreshness: A★ 2Claude Code +9 - 0.598score
pr-shepherd
nexttylabs
Use when a pull request needs one coordinated workflow covering review state, fixes, checks, branch updates, and merge readiness.
SkillAFreshness: A★ 2Claude Code +9 - 0.598score
respond-pr
nexttylabs
Use when a pull request already has review feedback that needs triage, code changes, verification, and clear replies back to reviewers.
SkillAFreshness: A★ 2Claude Code +9 - 0.598score
review-checklist
nexttylabs
Use when deciding whether a change is ready to merge and a concise cross-language checklist is more useful than a full narrative review.
SkillAFreshness: A★ 2Claude Code +9 - 0.598score
review-pr
nexttylabs
Use when a pull request needs a serious pre-merge review covering correctness, edge cases, security, tests, documentation, and architecture.
SkillAFreshness: A★ 2Claude Code +9
Done for you
Rather not set it up yourself?
Chalk Labs scopes, builds and runs agents for your team: the task audit, the eval harness, the integrations and the weekly ops report. Fixed scope, priced before we start.
See how we build AI agents →Read the case studies
Chalk Labs is a two-founder AI and growth studio. We build AI agents, CRMs and MVPs for Web3 and AI companies, and have shipped 50+ product and token launches.
- launches shipped
- 50+
- launches shipped
- media placements
- 500+
- media placements
- founders on every account
- 2
- founders on every account
Press we've placed: Forbes · CoinDesk · Decrypt · Cointelegraph · The Block
Rahil Jain, Co-founder, Chalk Labs. Ex-Goldman Sachs engineer. You talk to the person who builds it.
Get plan-mode running in your business
Tell us what you want to automate. A founder replies within one working day with a plan, not a pitch deck.
Cite this page
plan-mode skill. Agent Mart. https://agentmart.dev/skills/nexttylabs-plan-mode (last updated 2026-10-24)