PRAISONAIAGENTS · CVE-2026-40288 · AI agents & MCP
FIX → 1.5.140 · 4.5.139
PraisonAI's `praisonai workflow run` command executes the steps of a `type: job` workflow file, including `run:`, `script:` and `python:` steps, without checking what they contain. A malicious YAML file lets an attacker run arbitrary shell commands and Python code on the host that runs the workflow.
first public Apr 10, 2026 · 2 sources
Check it with your coding agent
It reads this page and tells you if your code or app is affected.
Teams that run praisonai workflow run on YAML files they did not write, including CI jobs that accept workflow files from pull requests and shared deployments where users supply workflows. Installations that only run their own trusted workflow files are exposed to less risk, but the flaw is in the code path that executes any type: job file.
Not affected if fixed-version, trusted-workflows-only
Who cares ai-agents, ci, praisonai, praisonaiagents, python
praisonai workflow run <file.yaml> loads untrusted YAML and if type: job executes steps through JobWorkflowExecutor in job_workflow.py.
This supports:
run: → shell command execution via subprocess.run()script: → inline Python execution via exec()python: → arbitrary Python script executionA malicious YAML file can execute arbitrary host commands.
action_run()_exec_shell(), _exec_inline_python(), _exec_python_script()Are you affected?
Ask from here. InfraStack checks the affected versions against your stack and answers in your inbox.
InfraStack checks the affected versions against your stack and answers in Slack, with the evidence and the fix.
InfraStack checks it against your stack and answers on your report, with the evidence and the fix.
pip show praisonai and pip show praisonaiagentspraisonai workflow run on workflow files from outside your team, such as CI pipelines that take YAML from pull requests or shared deploymentstype: job in workflow files you accept from untrusted sourcesUpgrade to 1.5.140 · 4.5.139
praisonai fixed in 4.5.139 (affected up to 4.5.138)praisonaiagents fixed in 1.5.140 (affected up to 1.5.139)Create exploit.yaml:
type: job
name: exploit
steps:
- name: write-file
run: python -c "open('pwned.txt','w').write('owned')"
Run:
praisonai workflow run exploit.yaml
exploit.yaml.praisonai workflow run exploit.yaml.pwned.txt appears in the working directory.Remote or local attacker-supplied workflow YAML can execute arbitrary host commands and code, enabling full system compromise in CI or shared deployment contexts.
Reporter: Lakshmikanthan K (letchupkt)
This page as .md · .json · CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:H/A:H
GitHub Advisory Database, CC-BY 4.0 · NVD data is provided by NIST and is public domain. Posts quoted belong to their authors.