agentmods.dev

autonomous-workshop: Skill for Claude Code

.claude/skills/build-a-toy/SKILL.md

build-a-toy is a skill for Claude Code from autonomous-ai/autonomous-workshop. It costs 122 tokens per session (8,764 once invoked), scanned A, original, Apache-2.0.

A skill for building a toy from a Design Contract, a written description of what the toy must be. It checks the build against every requirement and fixes differences until the result conforms or reports why it cannot.

In plain words
What is it for?
Use it after a design contract exists, or when an existing toy no longer matches its contract. It builds, checks, corrects, and publishes only a conforming result.
Why use it?
It prevents a build from being published while it still differs from the agreed design. The repeated review process makes missing or changed requirements visible.

Skill for Claude Code

Written for Claude Code: installed under .claude/. Also seen: reads .claude/ paths; mentions subagents; mentions Claude Code.

This is autonomous-ai/autonomous-workshop's own configuration. It tells Claude Code how to work on autonomous-workshop 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 autonomous-workshop configures →

not rated 24repo +1 changed today A scan 122 tokens original Apache-2.0
View source ↗ autonomous-ai/autonomous-workshop
Reuse

Borrowing it

Nothing to install: this file belongs to autonomous-ai/autonomous-workshop. 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/autonomous-ai/autonomous-workshop/main/.claude/skills/build-a-toy/SKILL.md
Clone the repo
git clone --depth 1 https://github.com/autonomous-ai/autonomous-workshop

Made for: Claude Code.

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 build-a-toy

README.md
[![agentmods](https://agentmods.dev/badge/skills/autonomous-ai/autonomous-workshop/build-a-toy/github.svg)](https://agentmods.dev/skills/autonomous-ai/autonomous-workshop/build-a-toy)
Your own site
<a href="https://agentmods.dev/skills/autonomous-ai/autonomous-workshop/build-a-toy"><img src="https://agentmods.dev/badge/skills/autonomous-ai/autonomous-workshop/build-a-toy/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 build-a-toy

Your own site · 80×15
<a href="https://agentmods.dev/skills/autonomous-ai/autonomous-workshop/build-a-toy"><img src="https://agentmods.dev/badge/skills/autonomous-ai/autonomous-workshop/build-a-toy.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 122 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 8,764 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.
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 $0.00122 $0.08764
Opus 5.5 $0.00049 $0.03506
Sonnet 5.5 $0.00024 $0.01753
Haiku 4.5 $0.00012 $0.00876

Measured today against content hash 095961668596, method: parsed. Prices are Anthropic first-party input rates as of 2026-10-07, from the pricing page.

Security

Grade A, and why

build-a-toy 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 today.

The scan reads SKILL.md. This mod also ships 7 executable files (scripts/ledger.py, scripts/runpaths.py, scripts/snapshot.py, …), listed below but not scanned — reading those needs a real analyzer, not pattern matching.

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.

.claude/skills/build-a-toy/SKILL.md · 589 lines

How it starts

The opening of the file, as written. The whole thing — 589 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Build a toy to its Design Contract

design-a-toy settles what the toy is. This skill owns everything after that: it builds the toy, checks the build against the contract, and corrects it until the two agree. The output is one published toy that conforms to its contract, or a plain report of why no build conformed.

Use the vocabulary in CONTEXT.md: Design Contract, Conformance, Requirement Scope, Unique Geometry, Component (not "part").

The settled run parameters

Every run this skill starts, whether wish or fix, uses:

Flag Value Why
--agent claude, on every wish and fix The CLI defaults to Codex, whose sandbox failed to start on the owner's build machine, and the agent is frozen for the run, so a wrong one costs a relaunch
--contract always Seals the Design Contract into the run (ADR 0072). The host refuses to complete the run unless its review carries every one of the contract's requirements, in order, each visibly matching — conformance for requirements is established by the run itself, not by this skill re-deriving coverage from a Manager-written list
--no-publish always, round 0 included Nothing reaches Factory until it conforms
--turn-minutes 360 The maximum. The clock is not the limit
--max-tokens 100000000 This is the real backstop. With 360-minute turns and unlimited resume, a run stops when its tokens run out
--ref every contract reference, every run The assembly round scores every sealed reference that no current component round already scored (ADR 0072). A description of an image is no substitute for the image
--guide every contract Guide Image (guides[], schema 5), every run, after the --ref images Guide Images explain the design to the Manager and the workers and are never compared (ADR 0089); the contract refuses a run that leaves one out
--print-art every contract print art file (print.art, schema 6), every run, after the --guide images The art printed on the toy's Print Sheets (ADR 0091), never compared as a likeness target; the contract refuses a run that leaves one out
--check-motion true when the contract has any moving part, on every wish, fix and resume workshop resume reselects this option and defaults it to false. Broken God v01 was resumed without it, and its final verification skipped the motion check and shipped with its wings unverified
--effort, --model only if the person passed them Both are frozen for the run and cannot change on resume, so never raise them yourself

Read the full file on GitHub · 589 lines

Files

What ships with it

10 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.

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. today Changed · +16 lines 095961668596
  2. yesterday Changed · +3 lines 41eef2a90d83
  3. 2d ago Changed · +5 lines b9305caa066c
  4. 3d ago Changed · +19 lines 49f9c7ea76ea
  5. 7d ago Changed · +24 lines 3ad8dcef16c5
  6. 13d ago Changed · +10 lines 353823c45b74
  7. 16d ago Changed · +1 lines 8e358a0300ff
  8. 22d ago Changed · +55 lines 86da56e39ddb
  9. 29d ago Changed · +26 lines 8ea18053d783
  10. 1mo ago First seen · 430 lines · 122 tokens per session scan A d3d2a2b6402e

Subscribe to this mod's changes

build-a-toy is a skill published in the GitHub repository autonomous-ai/autonomous-workshop (24 stars, last pushed yesterday), licensed Apache-2.0. It adds 122 tokens to every session and 8,764 once invoked, about $0.0005 per session on Opus 5.5. 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 2026-09-25.

3d-printingDesign and UXPythonaiai-agentsgenerative-designtoys
Related

build-a-toy compared with skills from other repositories

build-a-toy next to 6 skills from other repositories
ModTypeStarsLast pushLicenceScanTokensWorks with
build-a-toyautonomous-ai/autonomous-workshopSkill24yesterdayApache-2.0A122Claude Code
cocoindex-diagramscocoindex-io/cocoindexSkill12kyesterdayApache-2.0A124Claude Code, Codex
cli-commandComposioHQ/composioSkill31kyesterdayMITA76Claude Code, Codex
dashboard-designmckinsey/vizroSkill3.8kyesterdayApache-2.0A71Claude Code
dstack-prototypingdstackai/dstackSkill2.3kyesterdayMPL-2.0A80Claude Code, Codex
designing-vizro-layoutsmckinsey/vizroSkill3.8kyesterdayApache-2.0A58Claude Code
diagram-designvstorm-co/pydantic-deepagentsSkill1.1k7d agoMITA16Claude Code, Codex