Sources checked: . FailSafe Research.
A container or sandbox label does not tell a buyer which host files an agent can reach. A development assistant may need a project directory while a build tool also mounts a package cache, credentials directory, or container socket. A filesystem review measures those concrete boundaries using marked test files. It should explain what the agent process can read, modify, persist, or hand to another tool, rather than treating successful isolation of one directory as a complete containment result.
Filesystem access ledger
Inventory the runtime identity, mounted volumes, working directory, writable paths, inherited environment, subprocess model, and persistent caches. Define the files required for the authorized task and the resources explicitly outside scope. Use a disposable environment with synthetic markers placed by the operator in allowed and denied locations. Record the expected persistence after a job ends and after a worker restarts. If a managed platform controls the sandbox, obtain its documented boundary and identify the parts the assessment can actually observe.
Your filesystem access ledger
Use this compact record in the review. Fill it with synthetic fixtures and the observed result, then retain the configuration with the finding.
| Mounted path | Agent identity | Allowed action |
|---|---|---|
| Approved control case | Record the relevant decision | Record the observed result |
| Denied boundary case | Record the expected refusal | Record any deviation and owner |
Read and write approved fixture paths
Execute an ordinary task that reads and changes an allowed fixture. Then use the approved harness to request harmless reads and writes at denied test locations. Check shared caches and artifacts between two isolated test jobs. Observe whether a subprocess or file preview tool has the same restrictions as the primary agent runtime. Test cleanup through the supported job lifecycle rather than trying an escape technique. This review can establish access-policy gaps without publishing exploit instructions, probing the host, or manipulating production credentials.
Map effective filesystem access
Record the runtime and image version, mount policy, process identity, requested operation, expected decision, and observed file state. Include the operator’s marker manifest in restricted evidence so a reviewer can distinguish a denied read from a missing file. If a control exists only in the client interface, inspect whether the worker can still reach the resource through its normal tool. Separate filesystem findings from network, compute, and kernel-isolation questions. A file-boundary assessment should not be marketed as a full proof of sandbox security.
Verify the sandbox after a change
Retest ordinary work and denied paths after narrowing the runtime privileges. Check that cleanup removes job-specific artifacts while preserving only explicitly permitted caches. New mounts and build features should trigger review because they can expand access without changing the model. OWASP’s container security guidance offers background for runtime hardening. Use the actual platform configuration as the acceptance basis, and keep lower-level sandbox escape research out of scope unless a separate authorized engagement is designed for that purpose.
Plan this assessment with FailSafe
Use a staging environment, synthetic records, approved test identities, and an agreed stop condition. Record the model, connector, policy, and application versions so a later result can be compared with the same configuration. Unavailable evidence should remain an explicit coverage gap.
Primary reference: OWASP security guidance. The scenarios and acceptance criteria here are editorial recommendations for adapting that background to an authorized assessment.
For a scoped review of these boundaries, explore FailSafe AI security testing and agentic penetration testing. Agree the specific tools, identities, evidence, and exclusions before purchasing an assessment.
Questions & answers
Questions about this topic
Direct answers to common questions about this page.
No. The effective boundary depends on privileges, mounts, host interfaces, network policy, credentials, and the runtime configuration.
Yes. Operator-placed harmless fixtures and normal read/write tools can demonstrate whether intended path boundaries hold.
Related Articles
A sandbox, a target network and sovereign inference: three boundaries to review
Why an isolated agent runtime, target-side egress restrictions and private model inference solve different containment problems in an agentic pentest....
AI-Suggested Dependencies: A Review Workflow Before Installation
Check package identity, maintenance, version choice, and installation effects before accepting a coding agent’s dependency suggestion....
Onamae Desktop Cloud breach: review the provider-to-guest boundary
GMO’s Onamae Desktop Cloud disclosures describe malicious programs on some servers and later data publication. Lessons for hosted workload security....
Put this guidance to work
Bring an AI workflow with its identities and permitted actions. FailSafe can help turn this boundary into an authorized test and a useful remediation record.
Startup offer: $999 USD for 250 credits and one authorized black-box run. Sign in to SWARM before Stripe checkout. Custom scope and sovereign deployments are separate.