CVE-2026-55529— PraisonAI: Origin validation bypass in MCP HTTP Stream transport allows browser-mediated unauthenticated tool execution on local MCP server
Quick assessment
Affected
MervinPraison PraisonAI
Exploitation
No confirmed in-the-wild exploitation; assess based on exposure
Recommended action
Check the vendor advisory and references for a fixed version. If immediate upgrade is impossible, restrict exposure and increase monitoring.
PraisonAI 是一个多智能体团队系统。在 praisonai 4.6.58 版本之前，MCP HTTP Stream 的 方法使用 进行来源验证，这导致攻击者控制的 localhost.evil.example HTTP 来源与 localhost 的允许列表匹配。在没有 API 密钥的情况下，恶意网页可以向本地 MCP 服务器提交工具调用请求，并执行已暴露的工具。此问题已在版本 4.6.58 中修复。
Possible ATT&CK Techniques 2 AI
T1059 · Command and Scripting Interpreter T1190 · Exploit Public-Facing ApplicationAffected Version Matrix 1
| Vendor | Product | Version Range | Status |
|---|---|---|---|
| MervinPraison | PraisonAI | < 4.6.58 |
affected |
Get alerts for future matching vulnerabilities
Log in to subscribe
I. Basic Information for CVE-2026-55529
Vulnerability Information
Have questions about the vulnerability? See if Shenlong's analysis helps!
View Shenlong Deep Dive ↗
Although we use advanced large model technology, its output may still contain inaccurate or outdated information.Shenlong tries to ensure data accuracy, but please verify and judge based on the actual situation.
Vulnerability Title
PraisonAI: Origin validation bypass in MCP HTTP Stream transport allows browser-mediated unauthenticated tool execution on local MCP server
Source: CVE Program / CVE List V5
Vulnerability Description
PraisonAI is a multi-agent teams system. Prior to praisonai 4.6.58, the MCP HTTP Stream _validate_origin method accepts request_origin.startswith(allowed), so the attacker-controlled localhost.evil.example HTTP origin matches the localhost allowlist. Without an API key, a malicious webpage can submit tools/call requests to the local MCP server and execute exposed tools. This issue is fixed in version 4.6.58.
Source: CVE Program / CVE List V5
CVSS Information
CVSS:3.1/AV:N/AC:H/PR:N/UI:R/S:C/C:H/I:L/A:N
Source: CVE Program / CVE List V5
Vulnerability Type
关键功能的认证机制缺失
Source: CVE Program / CVE List V5
Affected Products
| Vendor | Product | Affected Versions | CPE | Subscribe |
|---|---|---|---|---|
| MervinPraison | PraisonAI | < 4.6.58 | - |
|
II. Public POCs for CVE-2026-55529
| # | POC Description | Source Link | Shenlong Link |
|---|
AI-Generated POC
Premium
No public POC found.
Login to generate AI POCIII. Intelligence Information for CVE-2026-55529
请登录查看更多情报信息。
Patches & Fixes for CVE-2026-55529 (1)
-
🔗 Release PraisonAI v4.6.58 · MervinPraison/PraisonAI · GitHub x_refsource_MISC
Vendor Advisories for CVE-2026-55529 (1)
Same Patch Batch · MervinPraison · 2026-08-25 · 17 CVEs total
| CVE-2026-55536 | 9.1 CRITICAL | Browser Server WebSocket origin validation bypass via unanchored regex (patch bypass of CV |
| CVE-2026-55541 | 8.8 HIGH | PraisonAI: `--api-key` flag on `praisonai serve` is not properly enforced |
| CVE-2026-55534 | 8.6 HIGH | PraisonAI serve agents --api-key is ignored, allowing unauthenticated remote agent executi |
| CVE-2026-55539 | 8.6 HIGH | PraisonAI: [Auth Bypass] PraisonAI async Jobs API (`/api/v1/runs`) has no authentication — |
| CVE-2026-55526 | 8.5 HIGH | PraisonAI: SSRF protection bypass in `spider_tools._host_is_blocked()` via DNS-resolved ho |
| CVE-2026-55528 | 8.2 HIGH | praisonaiagents: AgentServer declares auth_token but never enforces it on any route (CWE-8 |
| CVE-2026-55533 | 8.2 HIGH | PraisonAI: Authentication fail-open in Recipe server allows unauthenticated access when AP |
| CVE-2026-55532 | 7.6 HIGH | PraisonAI: Origin-validation bypass (startswith prefix match) enables unauthenticated cros |
| CVE-2026-55525 | 7.5 HIGH | PraisonAI: SSRF via redirect-following in praisonaiagents web_crawl |
| CVE-2026-55538 | 7.3 HIGH | PraisonAI: [Auth Bypass] `praisonai serve agents --api-key` is silently ignored — agent-in |
| CVE-2026-55537 | 7.1 HIGH | PraisonAI: Webhook SSRF via DNS fail-open in `JobSubmitRequest.validate_webhook_url()` — b |
| CVE-2026-55527 | 7.1 HIGH | PraisonAI: Arbitrary file write via unsanitized `user_id` in `FileMemory.__init__()` — pat |
| CVE-2026-55540 | 7.1 HIGH | PraisonAI: [Path Traversal] agent tools escape the configured workspace via symlinks |
| CVE-2026-55535 | 6.8 MEDIUM | PraisonAI: Server-Side Request Forgery via DNS rebinding bypass in webhook_url validation |
| CVE-2026-55531 | 6.5 MEDIUM | PraisonAI: Unauthenticated unbounded session accumulation in the PraisonAI MCP HTTP server |
| CVE-2026-55530 | 6.1 MEDIUM | PraisonAI: ast_grep_rewrite rewrites arbitrary files without the @require_approval gate en |
IV. Related Vulnerabilities
Same product: PraisonAI
-
CVE-2026-55536
Browser Server WebSocket origin validation bypass via unanch -
CVE-2026-55532
PraisonAI: Origin-validation bypass (startswith prefix match -
CVE-2026-55533
PraisonAI: Authentication fail-open in Recipe server allows -
CVE-2026-55539
PraisonAI: [Auth Bypass] PraisonAI async Jobs API (`/api/v1/ -
CVE-2026-55527
PraisonAI: Arbitrary file write via unsanitized `user_id` in
Same vendor: MervinPraison
-
CVE-2026-55536
Browser Server WebSocket origin validation bypass via unanch -
CVE-2026-55532
PraisonAI: Origin-validation bypass (startswith prefix match -
CVE-2026-55533
PraisonAI: Authentication fail-open in Recipe server allows -
CVE-2026-55539
PraisonAI: [Auth Bypass] PraisonAI async Jobs API (`/api/v1/ -
CVE-2026-55527
PraisonAI: Arbitrary file write via unsanitized `user_id` in
Same weakness: CWE-306
-
CVE-2026-80234
CAYIN Technology｜CAYIN CMS-WS/CMS-SE - Missing Authenticati -
CVE-2026-65105
NVIDIA NemoClaw未认证访问漏洞 -
CVE-2026-55640
Nextcloud MCP Server: Unauthenticated `POST /webhooks/nextcl -
CVE-2026-55539
PraisonAI: [Auth Bypass] PraisonAI async Jobs API (`/api/v1/ -
CVE-2026-55538
PraisonAI: [Auth Bypass] `praisonai serve agents --api-key`
V. Comments for CVE-2026-55529
No comments yet