CVE-2026-60087

PraisonAI before 1.6.78 Tool Approval Cache Bypass

Published Jul 15, 2026 CVSS 4.0 · 6.9 Medium EPSS < 1% KEV · Not listed
Description
PraisonAI before 1.6.78 caches tool approval decisions by tool name only, allowing attackers to reuse initial approvals for subsequent calls with arbitrary arguments. Attackers can exploit this by obtaining approval for a benign operation and then executing dangerous file write operations with unreviewed parameters in the same session.

Analysis

AI-generated

Technical impact

PraisonAI versions prior to 1.6.78 cache tool approval decisions based solely on tool name, allowing a previously approved non‑destructive tool to be invoked later with arbitrary arguments without re‑verification. An attacker who has first obtained approval for a benign operation can subsequently execute dangerous file write operations using unreviewed parameters within the same session, thereby gaining unauthorized privileged actions.

Affected Systems

All installations of MervinPraison PraisonAI earlier than version 1.6.78 are affected. Any deployment running those affected releases is vulnerable to the approval cache bypass and the associated privilege escalation.

Risk and Exploitability

The CVSS score of 6.9 indicates moderate severity. The EPSS score of less than 1% suggests the likelihood of widespread exploitation is low, and the vulnerability is not listed in the CISA KEV catalog. Exploitation requires a session where the attacker has already obtained approval for a benign tool; within that session the attacker can reuse the cached approval to invoke the tool with malicious arguments.

Remediation

No solution or workaround provided in the CVE record.

OpenCVE Recommended Actions

  • Upgrade PraisonAI to version 1.6.78 or later to apply the vendor fix.
  • If an immediate upgrade is not practical, disable or restrict tool approval caching so each invocation requires independent approval.
  • Enforce strict input validation for tool arguments, limit file‑write permissions, and monitor logs for anomalous file operations within sessions.

Generated by OpenCVE AI on July 31, 2026 at 03:59 UTC.

Tracking

Sign in to view the affected projects.

Advisories

Source ID Title
Github GHSA Github GHSA GHSA-29r9-67vg-qj56 PraisonAI: Human-in-the-loop tool approval is cached by tool name and silently reused for all subsequent calls with arbitrary arguments
History
Type Values Removed Values Added
First Time appeared Mervinpraison
Mervinpraison praisonai
Vendors & Products Mervinpraison
Mervinpraison praisonai
Type Values Removed Values Added
Metrics ssvc

{'options': {'Automatable': 'no', 'Exploitation': 'poc', 'Technical Impact': 'partial'}, 'version': '2.0.3'}

Type Values Removed Values Added
Description PraisonAI before 1.6.78 caches tool approval decisions by tool name only, allowing attackers to reuse initial approvals for subsequent calls with arbitrary arguments. Attackers can exploit this by obtaining approval for a benign operation and then executing dangerous file write operations with unreviewed parameters in the same session.
Title PraisonAI before 1.6.78 Tool Approval Cache Bypass
First Time appeared Praison
Praison praisonai
Weaknesses CWE-863
CPEs cpe:2.3:a:praison:praisonai:*:*:*:*:*:*:*:*
Vendors & Products Praison
Praison praisonai
References
Metrics cvssV3_1

{'score': 6.1, 'vector': 'CVSS:3.1/AV:L/AC:L/PR:N/UI:R/S:U/C:N/I:H/A:L'}

cvssV4_0

{'score': 6.9, 'vector': 'CVSS:4.0/AV:L/AC:L/AT:N/PR:N/UI:P/VC:N/VI:H/VA:L/SC:N/SI:N/SA:N'}

Weaknesses

Sources

Status
PUBLISHED
Assigner
VulnCheck
Published
Updated
Jul 15, 2026, 1:48 PM
Reserved
Jul 8, 2026, 12:14 PM
Status
Deferred
Published
Jul 15, 2026, 12:18 PM
Modified
Oct 24, 2026, 9:10 PM