Prompt Injection
231 entries
Every Prompt Injection entry VulnWatch has indexed, sorted by publication date.
Subscribe to this tag's RSS feed
CVE-2026-108600: open-multi-agent (@open-multi-agent/core) 1.5.0 through 1.21.2 contains a link following vulnerability in the file_write
open-multi-agent (@open-multi-agent/core) 1.5.0 through 1.21.2 contains a link following vulnerability in the file_write tool sandbox that allows attackers to create files outside the workspace root b...
CVE-2026-108599: phi 0.1.1 through 0.28.4 contains an improper link resolution vulnerability that allows malicious repositories to bypass
phi 0.1.1 through 0.28.4 contains an improper link resolution vulnerability that allows malicious repositories to bypass workspace_only_writes by exploiting lexical-only path checks in the permission...
CVE-2026-108592: mini-swe-agent 1.10.0 through 2.4.6 contains an information exposure vulnerability in BubblewrapEnvironment because bwra
mini-swe-agent 1.10.0 through 2.4.6 contains an information exposure vulnerability in BubblewrapEnvironment because bwrap omits --clearenv, so sandboxed commands inherit the host environment. Attacker...
CVE-2026-108583: zotero-mcp 0.10.0 through 0.14.1 contains a server-side request forgery vulnerability that allows attackers to reach int
zotero-mcp 0.10.0 through 0.14.1 contains a server-side request forgery vulnerability that allows attackers to reach internal services because _fetch_embedded_metadata fetches URLs without destination...
PraisonAI: Shell command allowlist bypass via find -exec built-in action
### Summary The shell command execution hardening introduced in PraisonAI npm 1.7.2 / Python 4.6.58 to fix GHSA-5jv7-2mjm-h6qj (utility-tools shell chaining) and GHSA-vjv9-7m7j-h833 (SandboxExecutor...
PraisonAI: AICoder Arbitrary File Write and Command Execution via LLM Tool Calls
### Summary The `AICoder` UI component exposes `write_to_file` and `execute_command` tools to the LLM with no path validation and no command sanitization. An attacker can achieve arbitrary file write...
PraisonAI: Prompt-injection defense blocks only when 3+ detector families fire simultaneously; realistic single-vector injections pass through unblocked
## Summary PraisonAI's opt-in prompt-injection defense (`enable_injection_defense()`) only blocks at `ThreatLevel.CRITICAL`, which requires three or more distinct detector families to match simultane...
PraisonAI: CodeAgent Executes LLM-Generated Code Without Sandboxing and Leaks All Environment Secrets
### Summary `CodeAgent._execute_python()` executes LLM-generated Python code in a subprocess with the complete parent-process environment (`os.environ.copy()`), zero AST validation, zero import restri...
PraisonAI: Prompt Injection Defense Bypassed for HIGH-Level Threats
### Summary Injection Defense is a purpose-built prompt injection defense layer. It scans LLM inputs through six detection categories: instruction overrides, authority claims, boundary manipulation, o...
Flowise Prompt Injection to RCE and SSRF via CSV/Airtable Agent Python Validator Bypass
## Summary Flowise
PraisonAI: AgentMail webhook lacks signature verification, allowing unauthenticated message injection and sender spoofing
## Summary PraisonAI's AgentMail bot, when run in webhook (or hybrid) mode, starts an aiohttp webhook server bound to `0.0.0.0` and processes inbound `message.received` events **without verifying any...
Langflow: Prompt injection in Langflow Smart Transform can lead to code execution
## Summary Langflow versions 1.3.0 through 1.10.2 contain a code-injection vulnerability in the Smart Transform (`LambdaFilterComponent`) component. Smart Transform places flow-author instructions a...
SiYuan MCP asset.upload Reads Arbitrary Absolute File Paths (Workspace Boundary Bypass)
# Security Advisory — SiYuan MCP `asset.upload` Reads Arbitrary Absolute File Paths (Workspace Boundary Bypass) | Field | Value | |---|---| | **Disclosed by** | joysinleung (`[email protected]`)...
Vibe-Trading LLM-callable tools permit command execution, code injection, and SSRF
### Summary: 5 findings — `BashTool` shell-injection sink (F6, the canonical RCE primitive), `BackgroundRunTool` async shell-injection sink (F7), backtest `exec_module()` runs top-level statements be...
CVE-2025-71427: Office-PowerPoint-MCP-Server through 2.0.7 contains a path traversal vulnerability that allows MCP callers to write and
Office-PowerPoint-MCP-Server through 2.0.7 contains a path traversal vulnerability that allows MCP callers to write and read files outside the working directory by supplying absolute paths or ../ sequ...
CVE-2026-96561: The AI Engine – The Chatbot, AI Framework & MCP for WordPress plugin for WordPress is vulnerable to Stored Cross-Site Sc
The AI Engine – The Chatbot, AI Framework & MCP for WordPress plugin for WordPress is vulnerable to Stored Cross-Site Scripting in versions up to, and including, 3.8.0 This is due to a chain of missin...
CVE-2026-102697: Ollama versions 0.14.0 before 0.31.2 contain an incorrect authorization vulnerability in the experimental agent mode Bas
Ollama versions 0.14.0 before 0.31.2 contain an incorrect authorization vulnerability in the experimental agent mode Bash tool approval mechanism that fails to properly parse shell syntax. Attackers w...
CVE-2026-77177: Open GenAI Stack (aka ogx-ai) 2026-06-11, as used in the Meta AI backend for WhatsApp and other products, allows code ex
Open GenAI Stack (aka ogx-ai) 2026-06-11, as used in the Meta AI backend for WhatsApp and other products, allows code execution because prompt injection (with Jinja2 template syntax) can be used to ac...
CVE-2026-97228: Rapid7 Bulk Export MCP versions 0.2.5 through 0.6.1 suffer from a GraphQL query injection issue in the export-status com
Rapid7 Bulk Export MCP versions 0.2.5 through 0.6.1 suffer from a GraphQL query injection issue in the export-status component (`get_export_status` in `src/export_manager.py`), whereby the `export_id`...
@bytebase/dbhub's read-only mode does not prevent database writes
### Summary Setting `readonly = true` on the `execute_sql` tool does not make the connection read-only. The connectors are written to set PostgreSQL `default_transaction_read_only=on` (and open SQLi...
DBHub HTTP transport DNS rebinding allows unauthenticated browser-origin SQL execution
### Summary DBHub `0.21.2` exposes an unauthenticated HTTP MCP endpoint when started with the documented HTTP transport mode, for example `--transport http --port 8080`. The HTTP server attempts to...
Decepticon: Role-boundary forgery via ChatML special-token literals in web crawl output composed into LLM context
## Summary Decepticon wraps web crawl results — the output of agent reconnaissance against target services — into LLM messages without neutralizing ChatML special-token literals. Under the BYOK (Brin...
CVE-2026-61742: DBHub is a database MCP server for Postgres, MySQL, SQL Server, Oracle, MariaDB, SQLite. Versions prior to 0.22.5 expose
DBHub is a database MCP server for Postgres, MySQL, SQL Server, Oracle, MariaDB, SQLite. Versions prior to 0.22.5 expose an unauthenticated HTTP MCP endpoint when started with the documented HTTP tran...
MCP Atlassian: Arbitrary file read/exfiltration via upload_attachment missing validate_safe_path()
## Summary The `upload_attachment` method in `confluence/attachments.py` reads and uploads arbitrary local files to Confluence without calling `validate_safe_path()`. Both download methods (`download...
MCP Atlassian: Arbitrary server-local file upload to Jira/Confluence attachments via unrestricted file_path parameters
## Summary The Jira and Confluence attachment upload tools accept caller-controlled file path parameters and read those paths from the MCP server's local filesystem before uploading the file as an At...