Executive brief
PraisonAI Platform, a system for managing multi-agent AI teams, contains a security flaw in how it handles labels and issues within workspaces. An authenticated user can modify, delete, or reassign labels and issues belonging to other users' workspaces by manipulating ID numbers in web requests. This could lead to unauthorized data modification, loss of organizational labels, and corruption of project tracking across different customer accounts.
Technical details
An Insecure Direct Object Reference (IDOR) exists in the PraisonAI Platform's label management endpoints, including PATCH, DELETE, and POST operations for labels and issues. The root cause is that while the application verifies the requester is a member of the workspace ID provided in the URL, it fails to verify that the specific label_id or issue_id provided also belongs to that workspace. The LabelService and associated route handlers pass these IDs directly to database queries without workspace-level scoping. An authenticated attacker can exploit this to modify or delete labels, or link/unlink labels to issues across different workspaces, provided they can guess or obtain the target object IDs. The issue is fixed in version 0.1.4.
Affected products
- MervinPraison praisonai-platform < 0.1.4
Timeline
- 2026-05-19: patched: Fix merged in commit ef79b7a
- 2026-07-21: disclosed: CVE-2026-47414 published