Attack ChainAIDerived
Hypothetical attack flow derived from CVE metadata – not a confirmed exploit chain
-
AccessAttain MITM position on developer's network
-
DeliveryVictim initiates Roo-Code Claude OAuth flow
-
ExploitIntercept cleartext OAuth callback traffic
-
ExecutionExtract authorization code or access token
-
PersistReplay token to OAuth provider
-
ImpactAccess victim's AI provider account
Vulnerability Assessment
| Weakness |
CWE
OWASP Top 10:2025
A04:2025 - Cryptographic Failures
ASVS 5.0
V12 - Secure Communication Transport encryption
|
| Exploitation Conditions | Exploitation requires the attacker to occupy a network-adjacent or network-path position between the victim developer's machine and the OAuth provider endpoint (consistent with AC:H in the CVSS 4.0 vector). The victim must have Roo-Code installed and must initiate or be mid-flight in an OAuth authentication flow through the Claude Code integration - this is the trigger that causes the cleartext transmission. No authentication or privileges are required of the attacker (PR:N). No user interaction beyond the normal OAuth login action is required. Limiting factors include the high attack complexity (network interception position is not trivially obtained outside of compromised Wi-Fi, corporate proxies, or ARP spoofing), the limited confidentiality impact (OAuth token scope determines downstream damage), and the requirement that the user actively use the OAuth feature rather than a pre-authenticated session. |
| Impact | The CVSS 4.0 score of 6.3 (Medium) with vector AV:N/AC:H/AT:N/PR:N/UI:N/VC:L reflects a network-reachable flaw requiring high attack complexity - consistent with the description characterizing exploitability as difficult, likely demanding a man-in-the-middle network position. Confidentiality impact is limited to the vulnerable system (VC:L) with no integrity or availability consequences, and no scope change to subsequent systems. The E:P supplemental metric confirms a public proof-of-concept exists, elevating real-world risk above theoretical. However, the most significant compounding risk factor is the product's end-of-life status: the vendor has archived the repository and will not issue a patch, leaving all users permanently exposed. The disclosure context notes multiple issues were reported to the vendor, who declined to remediate, so defenders cannot expect upstream resolution. EPSS data was not provided in the input and cannot be inferred. The limited blast radius (OAuth token exposure, not remote code execution) and high complexity prevent this from being a top-tier emergency, but the permanent no-patch status and public POC make it a genuine priority for any organization still running Roo-Code. |
| Exploit ScenarioDerived | An attacker with a man-in-the-middle position on the network path between a developer's workstation and the OAuth provider intercepts the OAuth callback traffic from Roo-Code's oauth.ts handler. Because the token or authorization code is transmitted in cleartext, the attacker captures the OAuth credential without needing any authentication or special software beyond a network sniffer. The captured token can then be replayed to access the associated AI provider account or any resource scoped to that OAuth grant. The public POC on GitHub demonstrates this interception technique. |
Get an e-mail when this CVE is exploited, patched or re-scored. Free account.
Alert meAnalysisAIv1 – Aug 28, 2026
Cleartext transmission of OAuth credentials in RooCodeInc Roo-Code up to version 3.51.1 exposes sensitive OAuth tokens to interception via the OAuth callback handler in src/integrations/claude-code/oauth.ts. The vendor has officially archived the repository and explicitly discourages continued use, meaning no patch will be issued. A public proof-of-concept is available on GitHub, though high attack complexity (requiring a network interception position) constrains opportunistic exploitation.
RemediationAI
No vendor-released patch has been identified at time of analysis, and none is expected - the vendor has explicitly archived the project and ceased support. The primary recommended action is to uninstall Roo-Code immediately and migrate to an actively maintained alternative AI coding assistant. If immediate removal is operationally infeasible, the following compensating controls should be applied: restrict Roo-Code usage to environments with enforced TLS inspection to detect and block cleartext OAuth token transmission; disable the Claude Code OAuth integration feature within Roo-Code if the feature can be toggled off; and ensure network segments where developers run Roo-Code enforce HTTPS-only egress to prevent cleartext exfiltration. Note that these controls reduce but do not eliminate exposure, as the root cause is within the application itself. The VulDB submission at https://vuldb.com/submit/864287 and the public POC gist at https://gist.github.com/tchen200311/dcf3d36781e8246876cde7965fdda490 can be used to validate whether compensating controls are effective in a test environment.
Affected VersionsENISA
Affected Packages
Source: OSV (Open Source Vulnerabilities)
Affects your stack
Affects 1 package in euvd
Checking this CVE against your own stacks and servers is part of the Basic plan and up.
Technical ContextAI · show details
CWE-319 (Cleartext Transmission of Sensitive Information) describes a root cause where sensitive data - in this case OAuth tokens or authorization codes - is transmitted over an unencrypted or insufficiently protected channel. The vulnerable component is the OAuth callback integration in Roo-Code's Claude Code integration layer (src/integrations/claude-code/oauth.ts), indicating the product handled OAuth flows for authentication against an AI provider API. The affected CPE is cpe:2.3:a:roocodeinc:roo-code:*:*:*:*:*:*:*:*, covering all versions through 3.51.1. The cleartext exposure likely involves OAuth authorization codes or access tokens being sent over HTTP rather than HTTPS, or being logged/transmitted without TLS enforcement in the callback path, making them recoverable by a network-positioned attacker.
Know when Roo Code gets its next vulnerability
- Watch products and get an email when something new is published
- Check this CVE against your own stack and servers
- Exploit and PoC references, threat intelligence from 22 sources
No credit card · 7-day full trial
Severity by source
CVSS:4.0/AV:N/AC:H/AT:N/PR:N/UI:N/VC:L/VI:N/VA:N/SC:N/SI:N/SA:N/E:P
Network vector with high complexity for MITM; no privileges or interaction needed; only limited confidentiality impact from token exposure.
Primary rating from Vendor (VulDB).
CVSS VectorVendor: VulDB
Lifecycle Timeline
3DescriptionCVE.org
A vulnerability was detected in RooCodeInc Roo-Code up to 3.51.1. This vulnerability affects unknown code of the file src/integrations/claude-code/oauth.ts of the component OAuth Callback. The manipulation results in cleartext transmission of sensitive information. The attack may be performed from remote. A high complexity level is associated with this attack. It is stated that the exploitability is difficult. The exploit is now public and may be used. Multiple isses were reported to the vendor beforehand. They explain, that "they all apply to Roo Code, a project we no longer support - the repository was archived a while ago, and we don't encourage anyone to use it." This vulnerability only affects products that are no longer supported by the maintainer.
Threat Intelligence Overview
How this decision was reached
The engine reads each signal below and applies a fixed list of rules from top to bottom. The first rule that matches sets the recommended action – like a statement, every line is shown.
What every action level means
SSVC AssessmentENISA
Priority Score
Attack Technique
Information disclosure occurs when an application unintentionally exposes sensitive data that aids attackers in reconnaissance or directly compromises security.
Arbitrary command execution in Roo Code (the open-source AI coding agent) through version 3.54.0 lets attackers slip pas
Roo Code is an AI-powered autonomous coding agent that lives in users' editors. Rated critical severity (CVSS 9.9), this
Roo Code is an AI-powered autonomous coding agent. Prior to 3.22.6, if the victim had "Write" auto-approved, an attacker
Roo Code is an AI-powered autonomous coding agent. The project-specific MCP configuration for the Roo Code agent is stor
Command approval bypass in Roo Code through 3.54.0 allows unauthenticated remote attackers to execute shell commands tha
Command-approval bypass in Roo Code through 3.54.0 lets attacker-influenced input cause the agent to execute arbitrary s
Auto-approval command filtering in Roo Code through 3.54.0 can be bypassed, letting an attacker run commands that should
Command approval bypass in Roo Code through 3.54.0 lets an attacker get denied shell commands auto-approved and executed
Roo Code through 3.54.0 can execute shell commands that its command auto-approval allowlist is meant to deny, because ge
Path traversal in RooCodeInc Roo-Code through version 3.51.1 allows remote unauthenticated attackers to read and write f
Code injection in RooCodeInc Roo-Code through version 3.51.1 allows remote attackers to execute arbitrary code via the E
Code injection in RooCodeInc Roo-Code up to version 3.51.1 allows a remote low-privileged attacker to execute injected c
EUVD-2026-67279
GHSA-57j6-cfwj-h582
NVD
CVE.org
EPSS
Vulnrichment
VulDB
Exploit-DB