A security-configuration guide for Wingthing, a tool that runs coding agents in isolated sandboxes. It asks about the agent's needs and creates an egg.yaml file describing what the sandbox should allow.

It is for configuring and hardening agent isolation on macOS or Linux.

It helps match sandbox restrictions to actual needs, such as network access, SSH, cloud credentials, or package installation.

create-egg — Skill by ehrlich-b

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/ehrlich-b/wingthing/create-egg

Any agent

npx skills add ehrlich-b/wingthing --skill create-egg

Clone the repo

git clone --depth 1 https://github.com/ehrlich-b/wingthing

Made for: Claude Code, Codex.

Per session 43 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.

When invoked 3,534 The whole file, excluding the scripts and references it only reads on demand.

Security scan B 1 finding. 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.

Model	Per session	Once invoked
Fable 5	$0.00043	$0.03534
Opus 5	$0.00022	$0.01767
Sonnet 5	$0.00009	$0.00707
Haiku 4.5	$0.00004	$0.00353

Measured 2d ago against content hash a38e340d84dc, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade B, and why

create-egg scanned grade B with 1 finding 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.

Reaches for credential filesmediumPrivilege escalation

SSH keys, cloud credentials, git-credentials, .npmrc, /etc/shadow: reading these is how a config file becomes a credential leak.

- "deny:~/.netrc" # HTTP auth blocked

Downgraded: this mod is about security review, or the phrase is quoted, so it is likely naming the pattern rather than instructing it.

docs/skills/create-egg/SKILL.md · 330 lines

How it starts

The opening of the file, as written. The whole thing — 330 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Egg Sandbox Expert

You are an expert security consultant specializing in wingthing egg sandbox configuration. You have deep knowledge of the sandboxing internals on both macOS (Seatbelt/SBPL) and Linux (user namespaces, seccomp BPF, mount isolation). You help users create egg.yaml files that are appropriately locked down for their use case.

Your Role

Interview the user about their security profile before generating config
Have strong opinions on what should be locked down beyond defaults
Explain the threat model — what each rule actually protects against
Generate production-ready egg.yaml files with comments explaining each choice

Interview Questions

Before generating a config, ask about:

What agent? (claude, codex, cursor, ollama, gemini) — each has different network/env needs
What kind of work? (coding a web app, infrastructure/DevOps, data science, general exploration)
Does the agent need SSH access? (git clone over SSH, deploy to remote servers)
Does the agent need cloud credentials? (AWS, GCP, Docker, Kubernetes)
Does the agent need to install packages? (npm, pip, cargo, go modules)
Are there secrets in your home directory? (.env files, API keys in dotfiles, password databases)
Is this a shared machine? (other users, CI runner, production host)
How long do sessions run? (quick prompts vs overnight unattended)
Platform? (macOS vs Linux — resource limits only work on Linux)

The Defaults (What You Get Without egg.yaml)

# These are the built-in defaults — you get these for free
fs:
  - "ro:/"                    # host filesystem readable; HOME writes are narrowed below
  - "rw:./"                   # project CWD writable
  - "rw:~/.cache/"            # build caches
  - "rw:~/Library/Caches/"    # macOS build caches
  - "rw:~/go/pkg/mod/cache/"  # Go module cache
  - "deny:~/.ssh"             # SSH keys blocked
  - "ro:~/.ssh/known_hosts"   # preserved so SSH can verify hosts
  - "deny:~/.gnupg"           # GPG keys blocked
  - "deny:~/.aws"             # AWS credentials blocked
  - "deny:~/.docker"          # Docker config blocked
  - "deny:~/.kube"            # Kubernetes config blocked
  - "deny:~/.netrc"           # HTTP auth blocked
  - "deny:~/.bash_history"    # shell history blocked
  - "deny:~/.zsh_history"     # shell history blocked
  - "deny-write:./egg.yaml"   # can't modify its own sandbox config
env: [HOME, PATH, TERM, LANG, USER]
# network: (none — agent profile auto-drills what's needed)

Read the full file on GitHub · 330 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.

2d ago First seen · 330 lines · 43 tokens per session scan B a38e340d84dc

Subscribe to this mod's changes

create-egg is a skill published in the GitHub repository ehrlich-b/wingthing (52 stars, last pushed 3d ago), licensed MIT. It adds 43 tokens to every session and 3,534 once invoked, about $0.0002 per session on Opus 5. A static security scan graded it B with 1 finding (reaches for credential files). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.

Security Writing and docs

systematic-debugging

Use when encountering any bug, test failure, or unexpected behavior, before proposing fixes.

obra/superpowers · 21 tokens

brainstorming

You MUST use this before any creative work - creating features, building components, adding functionality, or modifying behavior. Explores user intent, requirements and design before implementation.

obra/superpowers · 37 tokens

chat-pet-sprite-creation

Use when creating or changing VS Code chat pet sprite art, sprite sheets, state animations, eye treatments, Stable/Insiders variants, or pet transitions under src/vs/workbench/contrib/chat/browser/widget/media/chatPet.

microsoft/vscode · 53 tokens

cpu-profile-analysis

Analyze V8/Chrome CPU profiles (.cpuprofile) and DevTools trace files (Trace-.json). Use when: profiling performance, investigating slow functions, comparing code paths, finding bottlenecks, analyzing timeToRequest, understanding call trees from sampling profiler data, analyzing layout/paint/rendering, investigating…

microsoft/vscode · 71 tokens

babysit-pr

Babysit a GitHub pull request after creation by continuously polling review comments, CI checks/workflow runs, and mergeability state until the PR is merged/closed or user help is required. Diagnose failures, retry likely flaky failures up to 3 times, auto-fix/push branch-related issues when appropriate, and keep…

openai/codex · 114 tokens

imagegen

Generate or edit raster images when the task benefits from AI-created bitmap visuals such as photos, illustrations, textures, sprites, mockups, or transparent-background cutouts. Use when Codex should create a brand-new image, transform an existing image, or derive visual variants from references, and the output…

openai/codex · 113 tokens

create-egg