loop-setup

The one-time setup session for a development workflow called the loop. It examines the project, checks which tools and services actually work, and creates the loop's project files.

In plain words
What is it for?
Run it once before planning or building to detect the project structure, test and build commands, available connections, and the few decisions that require the owner's input.
Why use it?
It prevents the workflow from being configured around guesses about the project's stack, tests, Git setup, browser, database, or payment sandbox.

Skill for Claude CodeCodex

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add skills/cbdreamer11/cb-loop-kit-claude-plugin/loop-setup
Any agent
npx skills add cbdreamer11/CB-loop-kit-claude-plugin --skill loop-setup
Clone the repo
git clone --depth 1 https://github.com/cbdreamer11/CB-loop-kit-claude-plugin

Made for: Claude Code, Codex.

Per session 65 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 1,669 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5 $0.00065 $0.01669
Opus 5 $0.00032 $0.00834
Sonnet 5 $0.00013 $0.00334
Haiku 4.5 $0.00006 $0.00167

Measured 2d ago against content hash 5829502521d1, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

loop-setup scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

skills/loop-setup/SKILL.md · 123 lines

How it starts

The opening of the file, as written. The whole thing — 123 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Session 0 — set up the loop

You are setting up a working method in someone else's project. Touch nothing until they have seen what you propose to write. Never invent a fact about their stack: check it.

1 · Read the project

Detect, don't ask, what you can find yourself: ls, git remote -v, git branch --show-current, cat .gitignore, and whichever of these exist — package.json scripts, Makefile, pyproject.toml, Cargo.toml, go.mod, docker-compose.yml, CI config, test config, migrations directory. Report in three lines what this project is and how it is built.

2 · Check the connections — live, not by asking

The method depends on being able to observe things. Find out what is actually available in this session, right now, and print a table of PASS / FAIL / ABSENT:

Need How to check
Version control git status works, and whether a remote exists
Build / tests run the project's build or test command once
Browser is a browser-driving tool available (in-app browser, Chrome extension, Playwright/Puppeteer in devDependencies)? If yes, open one page and confirm it renders
Database is there a DB MCP server, a CLI (psql, mysql, sqlite3), or an ORM with a dev database? If yes, run one harmless read query
Payments sandbox if the project charges money: is a test-mode key present in the environment (never print its value)
Deploy how does this project publish, and is that command something the owner runs, not the agent

For anything ABSENT, say in one line what will not be verifiable without it and what the cheapest substitute is. Do not silently accept a gap — an absent connection becomes a declared GAP in .loop/VERIFY.md, not a shrug.

3 · Decide which models run which role

Ask, do not guess: which is the strongest model you have access to, and do you have one you would rather keep for the expensive steps? Plans differ, and a managed install can restrict models outright.

Explain the shape in one breath, then write the answer into .loop/roles/*.json:

Read the full file on GitHub · 123 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 2d ago First seen · 123 lines · 65 tokens per session scan A 5829502521d1

Subscribe to this mod's changes

loop-setup is a skill published in the GitHub repository cbdreamer11/CB-loop-kit-claude-plugin (8 stars, last pushed 1mo ago), licensed MIT. It adds 65 tokens to every session and 1,669 once invoked, about $0.0003 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.