Over 30 safety vulnerabilities have been disclosed in numerous synthetic intelligence (AI)-powered Built-in Improvement Environments (IDEs) that mix immediate injection primitives with professional options to realize information exfiltration and distant code execution.
The safety shortcomings have been collectively named IDEsaster by safety researcher Ari Marzouk (MaccariTA). They have an effect on well-liked IDEs and extensions akin to Cursor, Windsurf, Kiro.dev, GitHub Copilot, Zed.dev, Roo Code, Junie, and Cline, amongst others. Of those, 24 have been assigned CVE identifiers.
“I believe the truth that a number of common assault chains affected each AI IDE examined is probably the most shocking discovering of this analysis,” Marzouk instructed The Hacker Information.
“All AI IDEs (and coding assistants that combine with them) successfully ignore the bottom software program (IDE) of their menace mannequin. They deal with their options as inherently protected as a result of they have been there for years. Nevertheless, when you add AI brokers that may act autonomously, the identical options might be weaponized into information exfiltration and RCE primitives.”
At its core, these points chain three totally different vectors which might be widespread to AI-driven IDEs –
- Bypass a big language mannequin’s (LLM) guardrails to hijack the context and carry out the attacker’s bidding (aka immediate injection)
- Carry out sure actions with out requiring any consumer interplay by way of an AI agent’s auto-approved instrument calls
- Set off an IDE’s professional options that permit an attacker to interrupt out of the safety boundary to leak delicate information or execute arbitrary instructions
The highlighted points are totally different from prior assault chains which have leveraged immediate injections at the side of susceptible instruments (or abusing professional instruments to carry out learn or write actions) to switch an AI agent’s configuration to realize code execution or different unintended habits.
What makes IDEsaster notable is that it takes immediate injection primitives and an agent’s instruments, utilizing them to activate professional options of the IDE to end in info leakage or command execution.
Context hijacking might be pulled off in myriad methods, together with by means of user-added context references that may take the type of pasted URLs or textual content with hidden characters that aren’t seen to the human eye, however might be parsed by the LLM. Alternatively, the context might be polluted through the use of a Mannequin Context Protocol (MCP) server by means of instrument poisoning or rug pulls, or when a professional MCP server parses attacker-controlled enter from an exterior supply.
A few of the recognized assaults made doable by the brand new exploit chain is as follows –
- CVE-2025-49150 (Cursor), CVE-2025-53097 (Roo Code), CVE-2025-58335 (JetBrains Junie), GitHub Copilot (no CVE), Kiro.dev (no CVE), and Claude Code (addressed with a safety warning) – Utilizing a immediate injection to learn a delicate file utilizing both a professional (“read_file”) or susceptible instrument (“search_files” or “search_project”) and writing a JSON file by way of a professional instrument (“write_file” or “edit_file)) with a distant JSON schema hosted on an attacker-controlled area, inflicting the information to be leaked when the IDE makes a GET request
- CVE-2025-53773 (GitHub Copilot), CVE-2025-54130 (Cursor), CVE-2025-53536 (Roo Code), CVE-2025-55012 (Zed.dev), and Claude Code (addressed with a safety warning) – Utilizing a immediate injection to edit IDE settings information (“.vscode/settings.json” or “.thought/workspace.xml”) to realize code execution by setting “php.validate.executablePath” or “PATH_TO_GIT” to the trail of an executable file containing malicious code
- CVE-2025-64660 (GitHub Copilot), CVE-2025-61590 (Cursor), and CVE-2025-58372 (Roo Code) – Utilizing a immediate injection to edit workspace configuration information (*.code-workspace) and override multi-root workspace settings to realize code execution
It is value noting that the final two examples hinge on an AI agent being configured to auto-approve file writes, which subsequently permits an attacker with the flexibility to affect prompts to trigger malicious workspace settings to be written. However provided that this habits is auto-approved by default for in-workspace information, it results in arbitrary code execution with none consumer interplay or the necessity to reopen the workspace.
With immediate injections and jailbreaks appearing as step one for the assault chain, Marzouk provides the next suggestions –
- Solely use AI IDEs (and AI brokers) with trusted initiatives and information. Malicious rule information, directions hidden inside supply code or different information (README), and even file names can turn out to be immediate injection vectors.
- Solely connect with trusted MCP servers and constantly monitor these servers for modifications (even a trusted server might be breached). Evaluation and perceive the information circulation of MCP instruments (e.g., a professional MCP instrument may pull info from attacker managed supply, akin to a GitHub PR)
- Manually evaluation sources you add (akin to by way of URLs) for hidden directions (feedback in HTML / css-hidden textual content / invisible unicode characters, and so on.)
Builders of AI brokers and AI IDEs are suggested to use the precept of least privilege to LLM instruments, decrease immediate injection vectors, harden the system immediate, use sandboxing to run instructions, carry out safety testing for path traversal, info leakage, and command injection.
The disclosure coincides with the invention of a number of vulnerabilities in AI coding instruments that would have a variety of impacts –
- A command injection flaw in OpenAI Codex CLI (CVE-2025-61260) that takes benefit of the truth that this system implicitly trusts instructions configured by way of MCP server entries and executes them at startup with out in search of a consumer’s permission. This might result in arbitrary command execution when a malicious actor can tamper with the repository’s “.env” and “./.codex/config.toml” information.
- An oblique immediate injection in Google Antigravity utilizing a poisoned internet supply that can be utilized to govern Gemini into harvesting credentials and delicate code from a consumer’s IDE and exfiltrating the data utilizing a browser subagent to browse to a malicious web site.
- A number of vulnerabilities in Google Antigravity that would end in information exfiltration and distant command execution by way of oblique immediate injections, in addition to leverage a malicious trusted workspace to embed a persistent backdoor to execute arbitrary code each time the appliance is launched sooner or later.
- A brand new class of vulnerability named PromptPwnd that targets AI brokers linked to susceptible GitHub Actions (or GitLab CI/CD pipelines) with immediate injections to trick them into executing built-in privileged instruments that result in info leak or code execution.
As agentic AI instruments have gotten more and more well-liked in enterprise environments, these findings reveal how AI instruments broaden the assault floor of improvement machines, typically by leveraging an LLM’s lack of ability to tell apart between directions supplied by a consumer to finish a process and content material that it could ingest from an exterior supply, which, in flip, can include an embedded malicious immediate.
“Any repository utilizing AI for concern triage, PR labeling, code solutions, or automated replies is vulnerable to immediate injection, command injection, secret exfiltration, repository compromise and upstream provide chain compromise,” Aikido researcher Rein Daelman stated.
Marzouk additionally stated the discoveries emphasised the significance of “Safe for AI,” which is a brand new paradigm that has been coined by the researcher to deal with safety challenges launched by AI options, thereby making certain that merchandise usually are not solely safe by default and safe by design, however are additionally conceived holding in thoughts how AI parts might be abused over time.
“That is one other instance of why the ‘Safe for AI’ precept is required,” Marzouk stated. “Connecting AI brokers to present purposes (in my case IDE, of their case GitHub Actions) creates new rising dangers.”
