Profiles & Limits

Rustbox offers two execution profiles, each tuned for different workloads. Judge profile is available in the open beta. Executor profile is available on request.

Judge mode

Use case: Competitive programming, coding assessments, interview platforms.

Code runs in a bare sandbox with just the language runtime. No network access, tight memory limits, fast time limits. Designed for evaluating short, deterministic programs against known test cases.

Language	Memory	CPU Time	Wall Time	Max PIDs
Python	256 MB	4s	7s	10
C	256 MB	8s	10s	8
C++	256 MB	8s	10s	8
Java	512 MB	8s	10s	1024
JavaScript	512 MB	8s	12s	16
TypeScript	512 MB	8s	12s	16
Go	256 MB	8s	10s	1024
Rust	256 MB	8s	15s	64

All languages: network disabled, max code size 64 KB, max stdin size 256 KB.

Typical use: submit a candidate’s solution with test case input, compare stdout against expected output. Verdicts (AC, RE, TLE, MLE, PLE, FSE) are determined by kernel evidence, not exit code guessing.

Executor mode

Use case: AI agents, LLM tool execution, code playgrounds, interactive notebooks.

Relaxed limits accommodate real-world workloads. Filtered network allows standard library downloads and API calls while blocking private IP ranges and cloud metadata endpoints.

Resource	Limit
Memory	2 GB
Wall time	60 seconds
CPU time	60 seconds
Max PIDs	256
Network	Filtered
Code size	64 KB
Stdin size	256 KB

Typical use: an AI agent generates Python code that imports pandas, processes a CSV from stdin, and returns analysis results via stdout.

Comparison

	Judge	Executor
Memory	256-512 MB (per language)	2 GB
Wall time	7-15s (per language)	60s
CPU time	4-8s (per language)	60s
PIDs	8-1024 (per language)	256
Network	Disabled	Filtered
Best for	Assessments, contests	Agents, playgrounds

Security

Both profiles run with identical isolation: 8 kernel layers, seccomp filtering, capability drop, credential isolation, NO_NEW_PRIVS. The only differences are resource limits and network policy.

There is no way to escalate from one profile to the other via the API. Your profile is determined by your API key, not by the request.

Verdicts

Both profiles return the same verdict set:

Verdict	Meaning
AC	Accepted - clean exit, code 0
RE	Runtime Error - non-zero exit or crash
TLE	Time Limit Exceeded - wall time exceeded
MLE	Memory Limit Exceeded - OOM killed by cgroup
PLE	Process Limit Exceeded - fork/thread bomb hit cgroup PID limit
FSE	File Size Exceeded - SIGXFSZ from RLIMIT_FSIZE
SIG	Signaled - killed by signal
IE	Internal Error - platform infrastructure failure

Every verdict is backed by kernel evidence (cgroup OOM events, signal delivery records, CPU accounting). The platform does not guess verdicts from exit codes.