claude-vibe-squad: Skill for Claude Code

.agents/skills/auto-scaffold/SKILL.md

auto-scaffold is a skill for Claude Code, Codex from mtarcure/claude-vibe-squad. It costs 59 tokens per session (1,117 once invoked), scanned A, original, MIT.

A handoff-file generator for a verified repository that is missing standard project documents. It can create an absent README, CHANGELOG, LICENSE, or agent-context file from the work that was actually completed.

In plain words
What is it for?
Use it during local delivery when one or more approved handoff files are missing. It prepares only those missing files and leaves existing files unchanged.
Why use it?
It fills common delivery gaps without replacing documents that already exist. It keeps the handoff tied to the repository's real output rather than assumed features.

Skill for Claude CodeCodex

Written for no agent in particular: nothing here depends on one. Also seen: mentions CLAUDE.md; installed under .agents/ (shared by several agents).

This is mtarcure/claude-vibe-squad's own configuration. It tells Claude Code and Codex how to work on claude-vibe-squad itself, so it is not a mod to install elsewhere. Copy it as a starting point and replace the rules that are about this project. Everything claude-vibe-squad configures →

Reuse

Borrowing it

Nothing to install: this file belongs to mtarcure/claude-vibe-squad. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.

Copy the file
curl -O https://raw.githubusercontent.com/mtarcure/claude-vibe-squad/main/.agents/skills/auto-scaffold/SKILL.md
Clone the repo
git clone --depth 1 https://github.com/mtarcure/claude-vibe-squad

Made for: Claude Code, Codex.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

agentmods badge for auto-scaffold

README.md
[![agentmods](https://agentmods.dev/badge/skills/mtarcure/claude-vibe-squad/auto-scaffold/github.svg)](https://agentmods.dev/skills/mtarcure/claude-vibe-squad/auto-scaffold)
Your own site
<a href="https://agentmods.dev/skills/mtarcure/claude-vibe-squad/auto-scaffold"><img src="https://agentmods.dev/badge/skills/mtarcure/claude-vibe-squad/auto-scaffold/github.svg" alt="Measured on agentmods" height="20"></a>

Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.

agentmods 80×15 button for auto-scaffold

Your own site · 80×15
<a href="https://agentmods.dev/skills/mtarcure/claude-vibe-squad/auto-scaffold"><img src="https://agentmods.dev/badge/skills/mtarcure/claude-vibe-squad/auto-scaffold.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 59 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 1,117 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe. Third-party audits
  • NVIDIA SkillSpector warn 7 Sept 2026
SkillSpector: 1 finding, up to high

These are SkillSpector’s own severities. On a checked sample its high-severity flags on skills were ~96% false positives — a documented command, a public API, a “never do X” rule — so we show them as a caution to read, not a verdict. Why →

  • high Memory Poisoning · line 17
    Skill manipulates agent memory, state, or stored context. Memory corruption can alter personality, override safety rules, or cause unpredictable behavior.
    Fix: Protect agent memory and state from modification by untrusted content. Use read-only memory for critical instructions and validate all state changes.
How audits are shown
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 5.1 $0.00059 $0.01117
Opus 5 $0.00030 $0.00558
Sonnet 5 $0.00012 $0.00223
Haiku 4.5 $0.00006 $0.00112

Measured 10d ago against content hash e1407d91d0f0, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-10, from the pricing page.

Security

Grade A, and why

auto-scaffold 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 10d 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.

.agents/skills/auto-scaffold/SKILL.md · 79 lines

How it starts

The opening of the file, as written. The whole thing — 79 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Auto-Scaffold

Generate the standard repository files at delivery so a project arrives usable rather than as a bare tree of source. This runs during Phase 6 “Local deliver” of the S0–S7 project lifecycle, after verification and before local handoff. The reciprocal missing-standard-files trigger lives in shared/modes/project.md; that mode contract decides when this skill is loaded.

This skill writes files. That makes it a gated action, and the gate is not a formality — see Approval below.

The overwrite rule

Never overwrite a file that already exists. Skip it, and report the skip.

This is the rule that matters most, because every file this skill generates is a file a human is likely to have written by hand and cared about. A generated README replacing a hand-written one destroys work that generally is not recoverable from the branch, and the loss is silent — a scaffold step that reports success while having replaced the project's documentation. Detect by checking each target before writing. Overwriting requires explicit operator instruction naming the specific file.

What gets generated

  • README.md — title, one-line description, what it does, how to run it, and the tech stack. Derive from the project scope and from what the repo actually contains, not from what the plan intended.
  • CHANGELOG.md — derived from the commit history of the work. Group by change type. Where there is no meaningful history, write a minimal initial entry rather than fabricating a release narrative.
  • LICENSE — only with an operator-confirmed choice. See Approval.
  • CLAUDE.md — the agent-context file, so a future session in that repo starts oriented: what the project is, its conventions, how to build and test it, and the constraints that are not visible in the code. Note that CLAUDE.md is this repo's convention; the source pattern's companion context files have no counterpart here and are not created.

Steps

  1. Confirm the stage and read the tree first. Establish the work is verified and delivery is the next step, then enumerate which of the target files already exist. This list decides everything that follows.
  2. Gather the source material. Project scope, tech stack, entry points, build and test commands, and the commit history for the change log. Derive from the repository as it stands — a scaffold describing an intended project rather than the delivered one is actively misleading.
  3. Obtain operator approval before writing. Present the exact list of files to be created and the files being skipped because they exist. See Approval.
  4. Write only the missing files. Each existing target is skipped and reported by name.
  5. Verify the instructions you just wrote. Run the build and test commands exactly as the README states them. A scaffold's commands are its only load-bearing claim, and an untested command block is the most common defect in a generated README — it is written from the plan rather than from a run.
  6. Report what was created, what was skipped and why, and what remains for a human. Placeholders left for the operator are listed explicitly, never left to be discovered.

Read the full file on GitHub · 79 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. 10d ago First seen · 79 lines · 59 tokens per session scan A e1407d91d0f0

Subscribe to this mod's changes

auto-scaffold is a skill published in the GitHub repository mtarcure/claude-vibe-squad (142 stars, last pushed yesterday), licensed MIT. It adds 59 tokens to every session and 1,117 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 2026-08-30.