CVE-2026-12045

A flaw was found in the pgAdmin 4 AI Assistant. An attacker with the ability to influence database content that the assistant reads can exploit a transaction bypass vulnerability through prompt injection. This allows the attacker to execute arbitrary SQL queries with the privileges of the pgAdmin...

SQL Injection

Overview pgadmin4 is a PostgreSQL Tools Affected versions of this package are vulnerable to SQL Injection via the executesqlquery function. An attacker can execute arbitrary SQL statements and potentially achieve remote code execution by injecting a multi-statement payload through prompt injectio...

EUVD-2026-37965

Read-only transaction bypass in the pgAdmin 4 AI Assistant allows an attacker who can influence database content that the assistant reads to execute arbitrary SQL with the privileges of the pgAdmin user's database role. The AI Assistant's executesqlquery tool runs LLM-generated SQL inside a BEGIN...

GHSA-X44P-GG67-52FC Duplicate Advisory: PraisonAI: Coarse-Grained Tool Approval Cache Bypasses Per-Invocation Consent for Shell Commands

Duplicate Advisory This advisory has been withdrawn because it is a duplicate of GHSA-ffp3-3562-8cv3. This link is maintained to preserve external references. Original Description PraisonAI before 1.5.128 caches tool approval decisions by tool name only, not by invocation arguments, allowing...

CVE-2026-12045

Read-only transaction bypass in the pgAdmin 4 AI Assistant allows an attacker who can influence database content that the assistant reads to execute arbitrary SQL with the privileges of the pgAdmin user's database role. The AI Assistant's executesqlquery tool runs LLM-generated SQL inside a BEGIN...

CVE-2026-12045: Improper Neutralization of Special Elements used in an SQL Command

Read-only transaction bypass in the pgAdmin 4 AI Assistant allows an attacker who can influence database content that the assistant reads to execute arbitrary SQL with the privileges of the pgAdmin user's database role. The AI Assistant's executesqlquery tool runs LLM-generated SQL inside a BEGIN...

Evaluating LLMs for Real-World Web Vulnerability Detection

Large Language Models LLMs have emerged as a promising tool for automated vulnerability detection, yet their effectiveness on web-specific vulnerabilities remains to be explored. This work benchmarks six frontier Claude Opus 4.6, Codex GPT-5.4, Gemini 3.1-pro-preview and open-weight models Qwen...

PT-2026-53772

PathFilter's deny-list glob patterns are anchored, so .git, .obsidian, and node modules were only blocked at the vault root. Nested copies inside the vault e.g. tools/cli/node modules/..., tools/somerepo/.git/config, a nested .obsidian/ were fully traversable via isAllowed/isAllowedForListing...

CVE-2026-12045 pgAdmin 4: AI Assistant read-only transaction bypass allows unauthorised writes and remote code execution

Read-only transaction bypass in the pgAdmin 4 AI Assistant allows an attacker who can influence database content that the assistant reads to execute arbitrary SQL with the privileges of the pgAdmin user's database role. The AI Assistant's executesqlquery tool runs LLM-generated SQL inside a BEGIN...

CVE-2026-12045 pgAdmin 4: AI Assistant read-only transaction bypass allows unauthorised writes and remote code execution

Read-only transaction bypass in the pgAdmin 4 AI Assistant allows an attacker who can influence database content that the assistant reads to execute arbitrary SQL with the privileges of the pgAdmin user's database role. The AI Assistant's executesqlquery tool runs LLM-generated SQL inside a BEGIN...

CVE-2026-12045 pgAdmin 4: AI Assistant read-only transaction bypass allows unauthorised writes and remote code execution

Read-only transaction bypass in the pgAdmin 4 AI Assistant allows an attacker who can influence database content that the assistant reads to execute arbitrary SQL with the privileges of the pgAdmin user's database role. The AI Assistant's executesqlquery tool runs LLM-generated SQL inside a BEGIN...

CVE-2026-12045

Read-only transaction bypass in the pgAdmin 4 AI Assistant allows an attacker who can influence database content that the assistant reads to execute arbitrary SQL with the privileges of the pgAdmin user's database role. The AI Assistant's executesqlquery tool runs LLM-generated SQL inside a BEGIN...

CVE-2026-12045

The CVE-2026-12045 affects pgAdmin 4 (from version 9.13 up to before 9.16) and concerns the AI Assistant read-only transaction bypass. A prompt-injection vulnerability allows an attacker who can influence content seen by the AI Assistant to craft LLM-generated SQL payloads that bypass the BEGIN T...

CVE-2026-56074 PraisonAI - Tool Approval Cache Bypass via Coarse-Grained Caching

PraisonAI before 1.5.128 caches tool approval decisions by tool name only, not by invocation arguments, allowing subsequent executecommand calls to bypass approval prompts. Attackers can exploit this by obtaining initial approval for a benign command, then silently exfiltrate API keys and...

CVE-2026-56074

PraxionAI before 1.5.128 caches tool approval decisions by tool name rather than invocation arguments, enabling bypass of approval prompts for subsequent execute_command calls. Attackers could obtain initial approval for a benign command and then exfiltrate API keys and credentials via later shel...

GHSA-4H5R-5JM8-JXJM gemini-mcp-tool vulnerable to OS command injection and @file exfiltration via prompt quoting (CVE-2026-0755)

Untrusted prompt input could reach the Gemini CLI @file parser, allowing read/exfiltration of arbitrary local files @/etc/passwd, @/.ssh/idrsa, @../../secret. On Windows, unquoted cmd.exe metacharacters could break out into OS command injection. Fix 1.1.6: removed the broken shell:false...

gemini-mcp-tool vulnerable to OS command injection and @file exfiltration via prompt quoting (CVE-2026-0755)

Untrusted prompt input could reach the Gemini CLI @file parser, allowing read/exfiltration of arbitrary local files @/etc/passwd, @/.ssh/idrsa, @../../secret. On Windows, unquoted cmd.exe metacharacters could break out into OS command injection. Fix 1.1.6: removed the broken shell:false...

Unsafe Dependency Resolution

Overview @theia/ai-code-completion is a Theia - AI Core Affected versions of this package are vulnerable to Unsafe Dependency Resolution via the automatic loading of .prompts/.prompttemplate files in a workspace. An attacker can manipulate the AI agent's system instructions by introducing crafted...

Unsafe Dependency Resolution

Overview @theia/ai-claude-code is a Theia - Claude Code Integration Affected versions of this package are vulnerable to Unsafe Dependency Resolution via the automatic loading of .prompts/.prompttemplate files in a workspace. An attacker can manipulate the AI agent's system instructions by...

Unsafe Dependency Resolution

Overview @theia/ai-chat-ui is a Theia - AI Chat UI Extension Affected versions of this package are vulnerable to Unsafe Dependency Resolution via the automatic loading of .prompts/.prompttemplate files in a workspace. An attacker can manipulate the AI agent's system instructions by introducing...