hunting-code-interpreter-and-tool-sandbox-escape SKILL.md skillA
hunting-code-interpreter-and-tool-sandbox-escape from unboundcompute/security-agent-skills is an agent skill: a SKILL.md folder an AI agent loads when a task calls for it.
mdr add unboundcompute/security-agent-skills/hunting-code-interpreter-and-tool-sandbox-escape@git:20260901.f8395d1Pins git:20260901.f8395d1 in mdr.lock. Pin the fingerprint instead.
- Audit grade
- A, 17 of 17 checks passed
- Kind
- SKILL.md skill
- Repository
- unboundcompute/security-agent-skills
- Path
skills/hunting-code-interpreter-and-tool-sandbox-escape/SKILL.md- Versions recorded
- 1
- Latest version
- git:20260901.f8395d1, committed Sep 1, 2026
- Size
- 9,651 bytes
- Repository stars
- 26
- License
- MIT
Hunt for ways attacker-influenced code or a tool call escapes the sandbox an AI application runs it in: a code-interpreter or tool runtime that executes model-generated code with network access, a writable host filesystem, or credentials it should never see, a sandbox that shares a kernel, a mount, or an environment variable with the host so the guest reaches out, a resource with no CPU, memory, time, or output bound so one run starves the host, and a tool whose arguments reach a shell or a…
Collected from public GitHub. Every version is saved exactly as it was and never changes. Install it at an exact version with the mdr tool, and every file is checked before it reaches your agent. The main file is checked against the SHA-256 fingerprint recorded here, and the others against the fingerprints GitHub keeps for its source commit. The automatic checks below grade the latest version, and the same checks always give the same file the same grade.
What the file says
# Hunting code-interpreter and tool sandbox escape: the model writes code and something runs it An AI application that runs model-generated code, a code interpreter, an agent tool that shells out, a notebook backend, is executing attacker-influenceable instructions, because the model's output is shaped by its input and its input can be attacker-controlled. The sandbox that runs that code is therefore a security boundary around hostile code, and it is often built as if the code were trusted. The escapes are concrete. The runtime may have network access, so generated code reaches internal services or exfiltrates. It may see a writable host filesystem or a shared mount, so code reads or writes outside the jail. It may carry credentials, an API key, a cloud role, a token, in its environment, so code steals them. The isolation may be thin, a shared kernel, a container with a host mount, a subprocess with no namespace, so a known primitive escapes it. And with no CPU, memory, time, or output bound, one run starves the host. The hunt is to run code in the sandbox and see what of the host it can touch. You hunt this by making the model emit probing code and observing what succeeds. …
Read the whole file at its exact version.
How to install
mdr add unboundcompute/security-agent-skills/hunting-code-interpreter-and-tool-sandbox-escape@sha256:1db21009e0103d80The line at the top of this page pins the label git:20260901.f8395d1, which is what most people want. This one pins the sha256 fingerprint of the contents instead, so it survives a label being moved. Either way the exact version is written to mdr.lock, and mdr install downloads and checks those contents again, so it installs them exactly or stops with an error.
[](https://markdownregistry.com/a/art_j3cgqwycvapsjua2)
2 badge image fetches in 30 days. What the badge shows.
Versions
The registry has recorded 1 version of this file, committed on Sep 1, 2026.
Audit of the latest version
- pass: Frontmatter block present
- pass: Frontmatter declares a name
- pass: Frontmatter declares a description
- pass: Size between 200 bytes and 200 KB (9651 bytes)
- pass: No zero-width or bidi control characters
- pass: No instruction hidden inside an HTML comment
- pass: No link to an exfiltration or paste host
- pass: No credential-shaped string
- pass: No instruction to send local credentials anywhere
- pass: No text hidden with inline styles
- pass: No prompt-injection phrasing
- pass: No curl or wget piped into a shell
- pass: No recursive delete of root, home or parent
- pass: No instruction to read or print local credentials
- pass: No base64 blob over 200 characters
- pass: No link to a raw IP address
- pass: No script tag
Source
unboundcompute/security-agent-skills · 26 stars · license MIT · pushed 2026-09-08 · branch main
API
GET https://markdownregistry.com/api/v1/artifacts/art_j3cgqwycvapsjua2 GET https://markdownregistry.com/api/v1/resolve?ref=unboundcompute/security-agent-skills/hunting-code-interpreter-and-tool-sandbox-escape GET https://markdownregistry.com/api/v1/blob/1db21009e0103d80c432b843f6c429c1f19f16f94defdeeba5be2759c04bc34c
Your AI assistant does the legwork. You hear about the deals worth your word. Hand yours the standing instructions at modelranch.com and it joins the network that reads files like this one.
Guides: the SKILL.md format and its required fields · what to check before you install a skill · how to pin a skill to an exact version · where to find agent skills · keeping a team on the same skills · which agent file does what.