PT-2026-53137 · Pypi · Praisonai

Published

2026-06-18

·

Updated

2026-06-18

CVSS v3.1

7.3

High

VectorAV:N/AC:L/PR:L/UI:R/S:U/C:H/I:H/A:N

PraisonAI Code agent tools fail open without a workspace boundary

Summary

PraisonAI Code's agent-compatible CODE TOOLS wrappers keep a global workspace root initialized to None. If an application uses CODE TOOLS, code read file, code search replace, or code apply diff before calling set workspace(), the wrappers pass workspace=None into lower-level helpers that only enforce path containment when a workspace is truthy. Absolute paths outside the intended project workspace are then read and modified.
The official examples correctly call set workspace() before CODE TOOLS, and this report does not claim configured workspaces are ineffective. The issue is the fail-open default. PraisonAI's security documentation describes workspace boundaries as the path-traversal protection mechanism, and the already-published Python API arbitrary file write advisory (GHSA-hvhp-v2gc-268q) was fixed by defaulting an unset workspace to os.getcwd(). The adjacent read and edit paths reached through CODE TOOLS still fail open.

Affected Components

  • Package: praisonai
  • Current upstream main tested: 2f9677abb2ea68eab864ee8b6a828fd0141612e1
  • Latest tested release: v4.6.57
  • Primary files:
  • src/praisonai/praisonai/code/agent tools.py
  • src/praisonai/praisonai/code/tools/read file.py
  • src/praisonai/praisonai/code/tools/search replace.py
  • src/praisonai/praisonai/code/tools/apply diff.py

Root Cause

agent tools.py initializes workspace root to None and passes it directly to lower-level helpers:
python
 workspace root: Optional[str] = None
...
result = read file(..., workspace= workspace root)
...
result = search replace(..., workspace= workspace root)
The lower-level helpers only enforce containment if workspace is set:
python
if workspace:
  if not is path within directory(abs path, workspace):
    return {"success": False, ...}
The already-hardened write file() path uses effective workspace = workspace or os.getcwd(). Current tests assert that write file(workspace=None) must stay inside the current working directory. The same fail-closed default is missing from read file, search replace, apply diff, and the agent wrappers that call them.

Local-Only Reproduction

Run:
bash
PYTHONPATH=/path/to/PraisonAI/src/praisonai:/path/to/PraisonAI/src/praisonai-agents 
 python poc code tools workspace bypass.py
Expected vulnerable result:
text
[poc] HIT: CODE TOOLS wrappers read and edit outside workspace when workspace is unset
The PoV creates a temporary workspace and a temporary file outside that workspace. With get workspace() == None, code read file() reads the outside file, code search replace() modifies it, and code apply diff() modifies it again. After set workspace(workspace), the same outside path is rejected by all three wrappers.
No external services, model providers, or network access are used.

Impact

If an application exposes PraisonAI Code's agent-compatible CODE TOOLS to an LLM before setting a workspace boundary, prompt-influenced tool calls can read and modify files outside the intended project workspace. The practical attack shape matches the existing PraisonAI prompt-content advisory pattern: untrusted content influences an agent that has been given file-editing tools.
Practical impacts include:
  • reading host secrets or local configuration files accessible to the process user;
  • modifying arbitrary existing files when the attacker can supply or infer matching content for code search replace or code apply diff;
  • using code read file to first learn file content and then code apply diff to produce an exact modification;
  • bypassing the advertised workspace-boundary security posture unless the embedding application remembered to call set workspace() first.
This issue does not claim set workspace() is ineffective. The control works when configured. The vulnerability is the fail-open default for the advertised agent-tool bundle and adjacent read/edit helpers.

Affected-Version Sweep

The same behavior was reproduced on:
  • current upstream main: 2f9677abb2ea68eab864ee8b6a828fd0141612e1
  • v4.6.57
  • v4.6.56
  • v4.6.10
  • v4.6.9
  • v4.5.128
  • v4.5.126
  • v3.9.26
  • v3.9.24

Suggested Fix

Recommended fix:
  1. Make every low-level file helper compute effective workspace = workspace or os.getcwd() before resolving paths.
  2. Make code read file, code list files, code apply diff, code search replace, and code execute command use os.getcwd() as the default workspace when workspace root is None.
  3. Keep allowing absolute paths only when they resolve inside the effective workspace.
  4. Add regression tests proving outside absolute paths are rejected before and after set workspace().
  5. Consider failing closed if CODE TOOLS is used before a workspace is configured, or log a warning when the default current working directory is used.

Disclosure Route

PraisonAI's official security documentation lists GitHub Security Advisories as the preferred reporting method and asks reports to include reproduction steps, affected versions, impact, and suggested fixes. The repository security policy page currently shows no configured SECURITY.md, but private vulnerability reporting is available.

Fix

Information Disclosure

Incorrect Authorization

Path traversal

Found an issue in the description? Have something to add? Feel free to write us 👾
dbugs@ptsecurity.com

Weakness Enumeration

CWE-200CWE-863CWE-22

Related Identifiers

GHSA-GCQ3-MFVH-3X25

Affected Products

Praisonai