Live survey304 models · 14 providers · 22,792 auditsAudited grill-with-docs · SAFE →What changedHow we verifyllms.txt

Agent skill · Audited from source

Orchestrator Worker

skill · Th0rgal/sandboxed.shLicence not declaredAudited 14 Oct 2026 · source 1e416760e55d

★ 530pushed today
✓SAFEGrade B · 89/100 · 14 Oct 2026 · 1e41676
530repo stars · whole Th0rgal/sandboxed.sh

Is it safe? Yes — SAFE, grade B (89/100). We read this skill's source at commit 1e416760e55d and found no filesystem, network or shell access and no hard-coded secrets. What it does: Worker skill for boss-spawned missions. Stay within scope, verify, and report blockers quickly.

Ask an AI about this pageGChatGPTCClaudePPerplexityGeGeminiXGrok
More
Cite

Cite this page

ModelsAtlas. “Orchestrator Worker (Th0rgal/sandboxed.sh)”. Retrieved 14 Oct 2026. https://modelsatlas.com/skills/th0rgal/orchestrator-worker

Trust audit

Audit 0.4.1 · source read 14 Oct 2026

Trust score89/100✓SAFE B
0 critical0 high0 medium0 low0 info

SAFE means nothing in the source contradicts what the skill says it does. Grade B is shared by 16,667 of 17,602 audited skills.

What the source does

  • FilesystemNone observedNo reads or writes in the source
  • NetworkNone observedNo outbound calls in the source
  • ShellNone observedNo commands run in the source
  • DependenciesPinnedNothing floats
  • SecretsNone foundNo keys in source

Five audit layers

LayerWhat we checkResultFindings
L0 ProvenanceRepo owner, licence, archive status, activityPASS0
L1 Static sourceSecrets, path traversal, supply chain, hidden codeNA0
L2 InstructionsPrompt injection, authority framing, hidden directivesPASS0
L3 InventoryThe files and declarations match the sourcePASS0
L4 SandboxRun in isolation and watch behaviourSKIPPED why0

Gates applied: no behavioural pass.

Findings (0)

No findings at any severity. We re-audit when the source changes.

Audit 0.4.1 · source sha 1e416760e55d · 14 Oct 2026Report an issue or request a re-scan

Provenance

Who published it, and what the verdict is pinned to

  • PublisherTh0rgal GitHub userOwner of the repository this skill is read from
  • Source folderTh0rgal/sandboxed.sh/skills/orchestrator-workerThe exact folder we audited
  • LicenceNot declaredAs the repository states it
  • Commit pinned for this verdict1e416760e55dA new audit runs when the source changes
  • Repository statusNot archivedPushed today
  • Sandboxed runNot runSkills are read, not executed; layer L4 is skipped

What it tells the agent

Full source on GitHub ↗

Its description: Worker skill for boss-spawned missions. Stay within scope, verify, and report blockers quickly.

---
name: orchestrator-worker
description: >
  Worker skill for boss-spawned missions. Stay within scope, verify, and report
  blockers quickly.
---

# Orchestrator Worker

You are a worker spawned by a boss mission. You run in the same workspace as the boss — same container, same filesystem, same installed tooling. Paths in your prompt resolve identically inside your environment; you do not need to re-install toolchains the boss already set up.

## Rules

1. Stay inside the assigned scope. Do not widen the task on your own.
2. Work only in the provided working directory or branch.
3. Do not modify files outside your scope unless the boss explicitly expands it.
4. Verify with the command from the prompt before finishing. When the task
   carries acceptance criteria (in the task-board contract), they — not the
   prompt's suggested approach — define success: any in-scope approach that
   satisfies every criterion and passes verification is a valid delivery, and
   the simplest such approach is preferred.
5. Do not report `DONE` unless the files on disk actually match your claimed result.
6. If the prompt is wrong, the task is impossible, or scope is insufficient, report that immediately instead of exploring unrelated work.
7. Be concise. Prefer changes, verification, and a short status over long explanation.
8. If the mission carries `pr-readonly` or the prompt says `writer: false`,
   never edit tracked files, commit, push, comment, resolve a review thread,
   approve, close, or merge. Git/gh mutation guards are expected; do not try to
   bypass them. Report a reproducible finding for the separate writer.

## Communication

The boss may send follow-up messages or retask you. Treat them as updated instructions and reprioritize immediately.

## Completion

When done, make the result easy to integrate:
- commit on your branch if you changed files
- include the verification result
- include the changed file paths
- report one of: `DONE`, `BLOCKED`, or `NOT_FEASIBLE`

PR certifiers must instead finish with exactly one of:

```text
VERDICT: CLEAN
VERDICT: BLOCKED
VERDICT: INFRA_BLOCKED
```

Excerpt, first 2,120 characters of SKILL.md, verbatim from the audited commit. It is text from someone else’s repository, quoted so you can judge it before your agent reads it.

Install

Pinned to the audited commit 1e41676

Download folder ↗
git clone https://github.com/Th0rgal/sandboxed.sh
git -C sandboxed.sh checkout 1e416760e55d

The author documents no install command, so this uses the repository directly. Our verdict covers commit 1e416760e55d only.

Works in

The author has not said which agents this skill supports, and we have not tested it in any.

Audit history

Every audit, every source change

  1. ✓SAFE B 89 / 100

    first audit · source 1e416760e55d

Questions about Orchestrator Worker

What does the Orchestrator Worker skill do?
All your AI providers, machines, and agent harnesses in one place. Formerly sandboxed.sh (now the backend to the Orb clients).
Is Orchestrator Worker safe to install?
The audit found nothing in the source that contradicts what it says it does, and graded it B (89/100). Grade A is held back for packages that have also passed a sandboxed behavioural run, which is why a clean skill reads B.
What can Orchestrator Worker access on my machine?
The audit observed no filesystem, network or shell use at all in its source.
How current is this page?
The grade is for one exact copy of the source (1e416760e55d), read on 2026-10-14. The repository is watched, and a new audit runs when it changes — this is the first audit.

Sources

Spotted something wrong? Report a correction · Methodology

Keep going in your AI

Opens with this page as the source. We add nothing else to the prompt.