New: 98,000+ skills indexed from GitHub repos with 100+ stars · every skill safety-checkedBrowse top skillsTop skills
Esc to closeBrowse all skills
Submit a skillBrowse skills
Submit a skillBrowse skills

Agent skill

New Plan

by dcodesdev in dcodesdev/LetterSpace

Use this when the user wants to create a new plan, start a new feature, begin a new project, or says 'new plan'.

MITAuto-check passedAgent Workflows

Install New Plan

skills CLI
$ npx skills add dcodesdev/LetterSpace --skill new-plan -a claude-code

Project install by default; add -g for ~/.claude/skills/.

GitHub CLI
$ gh skill install dcodesdev/LetterSpace new-plan --agent claude-code

Project scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).

Manual copy
$ git clone --depth 1 https://github.com/dcodesdev/LetterSpace.git skills-src && mkdir -p .claude/skills && cp -r skills-src/.claude/skills/new-plan .claude/skills/new-plan && rm -rf skills-src

Use ~/.claude/skills/ instead of .claude/skills for a personal install. The folder must contain SKILL.md.

Claude Code skills documentation · loads skills from .claude/skills/

Facts

Skill name
new-plan
GitHub stars
100
Token cost
~1.2k tokens
SKILL.md length
548 words
Files
2
Skills in repo
5
Repo updated
First seen
Licence
MIT

At a glance

Use this when the user wants to create a new plan, start a new feature, begin a new project, or says 'new plan'.

  • Wants to create a new plan
  • SKILL.md covers Overview, The Process and Key Principles
  • Runs Shell scripts from its folder; calls bash
  • Start a new feature

What it does

New Plan is an agent skill from dcodesdev/LetterSpace. Use this when the user wants to create a new plan, start a new feature, begin a new project, or says 'new plan'. This skill defines a plan, writes a SPEC.md and PROGRESS.md into a timestamped plans/ directory, and updates PROMPT.md frontmatter to point to the new plan so the Ralph loop picks it up.

Its SKILL.md is about 1.2k tokens, which your agent loads only when the skill is triggered. The skill folder holds 1 other file (for example `new-plan.sh`).

It sits in Agent Workflows, covering Autonomous loops. The repository describes itself as: Self-hosted open source newsletter platform for managing subscribers and sending campaigns. The licence is MIT.

When your agent uses it

  • Wants to create a new plan
  • Start a new feature
  • Begin a new project

Example prompts

  • “new plan”
  • “/new-plan”

Requirements

  • A Bash shell

What it can do on your machine

Read from SKILL.md and the folder at commit a06ec86. It shows what the files ask for, not the result of running them.

  • Tool permissions

    Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.

    From allowed-tools in the SKILL.md frontmatter.

  • Runs code

    Ships script files (Shell), which the agent can run.

    Shell commands in SKILL.md call:

    • bash

    From the folder's file list and the shell code blocks in SKILL.md.

  • Network

    No URLs in SKILL.md.

    From URLs in SKILL.md, links to its own repository left out.

  • Credentials

    Names no API keys, tokens, secrets or passwords.

    From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.

Context cost

New Plan loads about 1.2k tokens when it runs. Until then it costs about 77 tokens; SKILL.md has 548 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it
~77
When it runs · the whole SKILL.md, loaded when a task matches
~1.2k

Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.

Safety

Auto-check passed

The automated check found no risky patterns in SKILL.md.

Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); files beside SKILL.md are not scanned.

SKILL.md

The full file from dcodesdev/LetterSpace at commit a06ec86, republished under its MIT licence (© dcodesdev). 548 words, ~1,246 tokens.

Download SKILL.mdSave it as .claude/skills/new-plan/SKILL.md (or your agent's skills folder). This skill also uses 1 other file; get the full folder from GitHub.
name
new-plan
description
Use this when the user wants to create a new plan, start a new feature, begin a new project, or says 'new plan'. This skill defines a plan, writes a SPEC.md and PROGRESS.md into a timestamped plans/ directory, and updates PROMPT.md frontmatter to point to the new plan so the Ralph loop picks it up.

New Plan

Overview

Create a structured project plan by defining a plan, then write it to disk and wire it up to PROMPT.md.

The Process

The order of these steps matters. Updating PROMPT.md must always be the very last step — only after SPEC.md and PROGRESS.md are fully written. A running loop picks up the new plan the moment PROMPT.md points at it, and a half-written plan would start executing.

1. Understand what the user wants to build:

  • Work out the scope, constraints, and success criteria of the request
  • Ask about anything genuinely ambiguous that would change the shape of the plan — but only if necessary
  • Ask through whatever mechanism this harness gives you: a question tool if you have one, whatever ask block the host app renders, otherwise plain prose in your reply
  • Necessary means: two readings of the request lead to materially different plans, and no sensible default picks between them. Otherwise pick the default, record it under ## Constraints in the spec, and carry on
  • Keep it to a couple of decisions at most. Never interrogate the user, and never block on a question you could answer by reading the codebase
  • If nobody is there to answer — an unattended loop, say — state the assumption in the spec instead

2. Explore the codebase thoroughly:

  • Before writing any plan, explore the existing codebase well to understand the current architecture, patterns, conventions, and relevant code
  • Read key files, understand the project structure, and identify where changes will need to be made
  • This ensures the plan is grounded in reality and accounts for existing code, dependencies, and constraints
  • Use this understanding to write phases and tasks that are specific to the actual codebase, not generic

3. Scaffold the plan directory:

Run the script bundled with this skill, passing only a short kebab-case name for the plan (e.g. auth-system, api-refactor):

bash
bash .claude/skills/new-plan/new-plan.sh <kebab-case-name>

It stamps the current date/time onto the directory, creates ./plans/YYYY-MM-DD-HHMM-<name>/, and prints the created path. It creates the directory only — you write SPEC.md and PROGRESS.md into it yourself. Use that path for everything below — never hand-compute the timestamp yourself.

4. Fill in the SPEC.md:

Create ./plans/YYYY-MM-DD-HHMM-<name>/SPEC.md:

markdown
# <Plan Name>

## Goal

One paragraph describing what we're building and why.

## Phases

### Phase 1: <Name>

- [ ] Task 1
- [ ] Task 2

### Phase 2: <Name>

- [ ] Task 1
- [ ] Task 2

(as many phases as needed — keep them small and focused)
(if a phase has too many tasks, split it into sub-phases: 2.a, 2.b, 2.c, etc.)

### Phase N: Documentation (always the last phase)

- [ ] Update existing docs affected by this feature (`README.md`, `CLAUDE.md`, `./docs/`, bundled skill templates)
- [ ] Create a new doc at `./docs/<feature>.md` if no existing doc covers it
- [ ] Do both when the feature both changes existing behavior and adds something new

## Constraints

- Any technical constraints, dependencies, or requirements

## Success Criteria

- How do we know this is done?

5. Create the PROGRESS.md:

Create ./plans/YYYY-MM-DD-HHMM-<name>/PROGRESS.md alongside the spec:

markdown
# Progress

## Completed Phases

(none yet)

## Current Phase

Not started.

## Notes

6. Update PROMPT.md (always last):

Do this last, after SPEC.md and PROGRESS.md are complete — never before or in between. Then update the frontmatter in ./PROMPT.md to point to the new plan:

yaml
---
plan: plans/YYYY-MM-DD-HHMM-<name>
---

The loop will automatically pick up the new plan by watching todo.md for unticked tasks.

Show full SKILL.md (141 more words)Show less

Key Principles

  • One phase = one iteration — each phase should be completable in a single Ralph loop run
  • Split large phases — if a phase would have too many tasks (e.g. implementing 20 API routes), break it into sub-phases like Phase 2.a, 2.b, 2.c, 2.d with ~5 tasks each. Each sub-phase should still be completable in one iteration.
  • Small and concrete — tasks should be specific, not vague
  • Ordered by dependency — earlier phases should not depend on later ones
  • PROMPT.md last — never point PROMPT.md at the plan until every plan file is written
  • YAGNI — don't plan features that aren't needed yet
  • Docs last — the final phase of every plan MUST be a documentation phase. Update existing docs where they cover the touched behavior, create a new doc under ./docs/ when the feature isn't covered anywhere, or do both. Never end a plan without it.

© dcodesdev, MIT. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file

Files

SKILL.md and 1 other file in .claude/skills/new-plan of dcodesdev/LetterSpace.

  • SKILL.md
  • new-plan.sh

Open the folder on GitHubat commit a06ec86

Compare with similar skills

New Plan next to the 5 skills that share the most tags, products or categories with it. Stars are the repository's; “used in” counts other GitHub owners with a copy.

New Plan compared with similar skills
SkillStarsUsed inTokensAuto-checkLicenceRepo updated
New Plan this skilldcodesdev/LetterSpace100—~1.2kAutomated safety check: PassMIT
Show Me Your Work Decision Logcursor/plugins10k9 repos~1.6kAutomated safety check: PassNone
Autoresearch Iteration Loopuditgoenka/autoresearch6.5k1 repos~2kAutomated safety check: PassMIT
PUA Looptanweai/pua20k1 repos~1.1kAutomated safety check: PassMIT
AutopilotYeachan-Heo/oh-my-claudecode40k1 repos~4.4kAutomated safety check: PassMIT
Install Loop Engineeringcobusgreyling/loop-engineering11k1 repos~648Automated safety check: PassMIT

Similar skills

  • Show Me Your Work Decision Log

    cursor/plugins

    Official

    Keeps a TSV decision log for long or unattended agent runs, one row per decision with what, why, evidence and result, so a reviewer can check the work later.

    10k GitHub starsUsed in 9 repos~1.6k tokens
    Agent WorkflowsAuto-check passed

  • Autoresearch Iteration Loop

    uditgoenka/autoresearch

    Runs an autonomous modify, verify, keep-or-discard loop against any metric, with subcommands for planning, debugging, fixing, security audits, shipping and more.

    6.5k GitHub starsUsed in 1 repo~2k tokens
    Agent WorkflowsAuto-check passed

  • PUA Loop

    tanweai/pua

    Runs an unattended iterate-until-verified loop in which a user-set verify command, not the agent's own claim, decides when the task is finished.

    20k GitHub starsUsed in 1 repo~1.1k tokens
    Agent WorkflowsAuto-check passed

  • Autopilot

    Yeachan-Heo/oh-my-claudecode

    Takes a short product idea through requirements, design, planning, parallel implementation, QA cycles and multi-reviewer validation to produce working code.

    40k GitHub starsUsed in 1 repo~4.4k tokens
    Agent WorkflowsAuto-check passed

  • Install Loop Engineering

    cobusgreyling/loop-engineering

    Installs Loop Engineering into a project through the single @cobusgreyling/loop CLI, scaffolding a report-only loop and a readiness score.

    11k GitHub starsUsed in 1 repo~648 tokens
    Agent WorkflowsAuto-check passed

  • Loopy

    Forward-Future/loopy

    Discover, find, compare, audit, repair, adapt, craft, run, debrief, save, and prepare repeatable AI-agent loops for publication.

    3.2k GitHub stars~3.9k tokensUpdated 27 days ago
    Agent WorkflowsAuto-check passed

More from dcodesdev/LetterSpace

  • Design Brief

    dcodesdev/LetterSpace

    Use this when the user wants something designed — a screen, page, flow, app, or redesign — or says 'design', 'design brief', 'UI for'.

    100 GitHub stars~1.4k tokensUpdated 5 days ago
    Auto-check passed

  • Refactor

    dcodesdev/LetterSpace

    Use this when the user wants to refactor the codebase, clean up code, reduce duplication, or says 'refactor' / 'new refactor'.

    100 GitHub stars~1.4k tokensUpdated 5 days ago
    Auto-check passed

  • Document

    dcodesdev/LetterSpace

    Use this when writing or updating documentation — a README, docs/ page, CLAUDE.md, changelog entry, or doc comments — or when the user says 'document this'.

    100 GitHub stars~461 tokensUpdated 5 days ago
    Auto-check passed

  • Brainstorm

    dcodesdev/LetterSpace

    Use this when the user wants to brainstorm, think through an idea, explore approaches, weigh options, or says 'brainstorm' / 'let's think about'.

    100 GitHub stars~380 tokensUpdated 5 days ago
    Auto-check passed

Categories

Questions about New Plan

What does New Plan do?

Use this when the user wants to create a new plan, start a new feature, begin a new project, or says 'new plan'. New Plan is an agent skill from dcodesdev/LetterSpace. Use this when the user wants to create a new plan, start a new feature, begin a new project, or says 'new plan'.

When should I use New Plan?

New Plan fits situations like: wants to create a new plan; start a new feature; begin a new project.

How do I install New Plan in Claude Code?

Run `npx skills add dcodesdev/LetterSpace --skill new-plan -a claude-code`. Or copy the skill folder (.claude/skills/new-plan in dcodesdev/LetterSpace) into .claude/skills/new-plan in your project. Claude Code loads it when a task matches its description.

How do I install New Plan in Codex?

Run `npx skills add dcodesdev/LetterSpace --skill new-plan -a codex`. Or copy the skill folder (.claude/skills/new-plan in dcodesdev/LetterSpace) into .agents/skills/new-plan in your project. Codex loads it when a task matches its description.

Can I use New Plan in Cursor, Gemini CLI or GitHub Copilot?

Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add dcodesdev/LetterSpace --skill new-plan -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/new-plan, .gemini/skills/new-plan, .github/skills/new-plan and .opencode/skills/new-plan in your project.

What does New Plan need to run?

Going by SKILL.md and its folder, New Plan needs a shell for the scripts in its folder and the command-line tools its instructions call (bash). Our summary lists: A Bash shell.

Does New Plan access the network?

SKILL.md contains no URLs. Any network use would come from the scripts or tools the agent runs. This is read from the text; nothing was executed.

Is New Plan safe to install?

Our automated static check of SKILL.md found no risky patterns, such as piping downloads into a shell, reading credential files or hidden Unicode. It is not a guarantee. Review the folder before installing.

What licence does New Plan use?

New Plan is published under the MIT licence (the repository's licence). It allows redistribution, so the full SKILL.md is shown on this page.

How many tokens does New Plan use?

About 1.2k tokens (SKILL.md is roughly 5k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full.

What are the alternatives to New Plan?

Skills that share tags, products or a category with New Plan: Show Me Your Work Decision Log (cursor/plugins, 10k stars), Autoresearch Iteration Loop (uditgoenka/autoresearch, 6.5k stars), PUA Loop (tanweai/pua, 20k stars) and Autopilot (Yeachan-Heo/oh-my-claudecode, 40k stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.

Who maintains New Plan?

dcodesdev (a GitHub user) maintains it in dcodesdev/LetterSpace, which has 100 GitHub stars. The repository holds 5 skills in this directory. The repository was last updated on October 2, 2026.

Source: dcodesdev/LetterSpace on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.