CVE-2026-39891
published 2026-04-08CVE-2026-39891: PraisonAI is a multi-agent teams system. Prior to 4.5.115, the create_agent_centric_tools() function returns tools (like acp_create_file) that process file…
PriorityP352high8.8CVSS 3.1
AVNACLPRLUINSUCHIHAH
EPSS
0.56%
42.2th percentile
PraisonAI is a multi-agent teams system. Prior to 4.5.115, the create_agent_centric_tools() function returns tools (like acp_create_file) that process file content using template rendering. When user input from agent.start() is passed directly into these tools without escaping, template expressions in the input are executed rather than treated as literal text. This vulnerability is fixed in 4.5.115.
Affected
3 ranges
| Vendor | Product | Version range | Fixed in |
|---|---|---|---|
| mervinpraison | praisonai | < 4.5.115 | 4.5.115 |
| mervinpraison | praisonai | >= 0 < 4.5.115 | 4.5.115 |
| praison | praisonai | <= 4.5.114 | — |
Stop checking back — get the weekly exploitation signal.
Every Monday: what got weaponized or added to CISA KEV in the last seven days — each CVE cross-linked to its PoC, Nuclei template, and detection rule. Free, one email a week, unsubscribe in one click.
GHSA
PraisonAI has Template Injection in Agent Tool Definitions
ghsa·2026-04-08
CVE-2026-39891 [HIGH] CWE-94 PraisonAI has Template Injection in Agent Tool Definitions
PraisonAI has Template Injection in Agent Tool Definitions
## Summary
Direct insertion of unescaped user input into template-rendering tools allows arbitrary code execution via specially crafted agent instructions.
## Details
The `create_agent_centric_tools()` function returns tools (like `acp_create_file`) that process file content using template rendering. When user input from `agent.start()` is passed directly into these tools without escaping (as shown in `agent_centric_example.py:85-86`), template expressions in the input are executed rather than treated as literal text. This occurs because:
1. No input sanitization or escaping is applied to user-controlled content
2. The ACP-enabled runtime auto-approves operations (`approval_mode="auto"`)
3. Tools lack context-aware escaping for te
OSV
PraisonAI has Template Injection in Agent Tool Definitions
osv·2026-04-08
CVE-2026-39891 [HIGH] PraisonAI has Template Injection in Agent Tool Definitions
PraisonAI has Template Injection in Agent Tool Definitions
## Summary
Direct insertion of unescaped user input into template-rendering tools allows arbitrary code execution via specially crafted agent instructions.
## Details
The `create_agent_centric_tools()` function returns tools (like `acp_create_file`) that process file content using template rendering. When user input from `agent.start()` is passed directly into these tools without escaping (as shown in `agent_centric_example.py:85-86`), template expressions in the input are executed rather than treated as literal text. This occurs because:
1. No input sanitization or escaping is applied to user-controlled content
2. The ACP-enabled runtime auto-approves operations (`approval_mode="auto"`)
3. Tools lack context-aware escaping for te
No detection rules found.
No public exploits indexed.
2026-04-08
Published