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.
npx agentmods add skills/googlecloudplatform/scion/git-sandboxnpx skills add GoogleCloudPlatform/scion --skill git-sandboxgit clone --depth 1 https://github.com/GoogleCloudPlatform/scionWhat 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 | $0.00042 | $0.00562 |
| Opus 5 | $0.00021 | $0.00281 |
| Sonnet 5 | $0.00008 | $0.00112 |
| Haiku 4.5 | $0.00004 | $0.00056 |
Grade A, and why
git-sandbox 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.
How it starts
The opening of the file, as written. The whole thing — 47 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Git Workflow Protocol: Sandbox & Worktree Environment
You are operating in a restricted, non-interactive sandbox environment. Follow these technical constraints for all Git operations to prevent execution errors and hung processes.
1. Local-Only Operations (No Network Access)
- Restriction: The environment is air-gapped from
origin. Commands likegit fetch,git pull, orgit pushwill fail. - Directive: Always assume the local
mainbranch is the source of truth. - Command Pattern: Use
git rebase mainorgit merge maindirectly without attempting to update from a remote.
2. Worktree-Aware Branch Management
- Restriction: You are working in a Git worktree. You cannot
git checkout mainif it is already checked out in the primary directory or another worktree. - Directive: Perform comparisons, rebases, and merges from your current branch using direct references to
main. Do not attempt to switch branches to inspect code. - Reference Patterns:
- Comparison:
git diff main...HEAD(to see changes in your branch). - File Inspection:
git show main:path/to/file.ext(to view content on main without switching). - Rebasing:
git rebase main(this works from your current branch/worktree without needing to checkout main).
- Comparison:
3. Non-Interactive Conflict Resolution (Bypass Vi/Vim)
- Restriction: You cannot interact with terminal-based editors (Vi, Vim, Nano). Any command that triggers an editor will cause the process to hang.
- Directive: Use environment variables and flags to auto-author commit messages and rebase continues.
- Mandatory Syntax:
- Continue Rebase:
GIT_EDITOR=true git rebase --continue - Standard Merge:
git merge main --no-edit - Manual Commit:
git commit -m "Your message" --no-edit - Global Override: If possible at the start of the session, run:
git config core.editor true
- Continue Rebase:
4. Conflict Resolution Loop
If a rebase or merge results in conflicts:
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 · 47 lines · 42 tokens per session scan A c2726e53c06f
git-sandbox is a skill published in the GitHub repository GoogleCloudPlatform/scion (1,683 stars, last pushed 2d ago), licensed Apache-2.0. It adds 42 tokens to every session and 562 once invoked, about $0.0002 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.
Other skills, from other repositories
systematic-debugging
Use when encountering any bug, test failure, or unexpected behavior, before proposing fixes.
next-cache-components-adoption
Turn on Cache Components in a Next.js app and resolve the blocking routes it surfaces. Use when the user wants to enable, adopt, or migrate to Cache Components, flip the cacheComponents flag, work through a flood of blocking-prerender / instant validation errors, run the cache-components-instant-false codemod, or…
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…
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…
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…
next-cache-components-optimizer
Drive a Next.js route to instant navigation by setting up an agentic loop, under Cache Components / PPR, on initial load (hard navigation) and client-side navigation (soft navigation). Encode the goal as a failing @next/playwright instant() e2e and work it to green, one verified route at a time; the shipped test then…