Passing the Safety Vibe Test: The Risks of Vibe Coding

Introduction

At Databricks, our AI Purple Crew commonly explores how new software program paradigms can introduce sudden safety dangers. One current development we have been monitoring intently is “vibe coding”, the informal, speedy use of generative AI to scaffold code. Whereas this method accelerates growth, we have discovered that it may possibly additionally introduce refined, harmful vulnerabilities that go unnoticed till it is too late.

On this submit, we discover some real-world examples from our pink workforce efforts, exhibiting how vibe coding can result in severe vulnerabilities. We additionally show some methodologies for prompting practices that may assist mitigate these dangers.

Vibe Coding Gone Improper: Multiplayer Gaming

In considered one of our preliminary experiments exploring vibe coding dangers, we tasked Claude with making a third-person snake battle area, the place customers would management the snake from an overhead digicam perspective utilizing the mouse. In keeping with the vibe-coding methodology, we allowed the mannequin substantial management over the venture’s structure, incrementally prompting it to generate every part. Though the ensuing utility functioned as supposed, this course of inadvertently launched a crucial safety vulnerability that, if left unchecked, may have led to arbitrary code execution.

The Vulnerability

The community layer of the Snake recreation transmits Python objects serialized and deserialized utilizing pickle, a module recognized to be susceptible to arbitrary distant code execution (RCE). Because of this, a malicious consumer or server may craft and ship payloads that execute arbitrary code on another occasion of the sport.

The code under, taken immediately from Claude’s generated community code, clearly illustrates the issue: objects acquired from the community are immediately deserialized with none validation or safety checks.

Though such a vulnerability is basic and well-documented, the character of vibe coding makes it simple to miss potential dangers when the generated code seems to “simply work.”

Nonetheless, by prompting Claude to implement the code securely, we noticed that the mannequin proactively recognized and resolved the next safety points:

As proven within the code excerpt under, the difficulty was resolved by switching from pickle to JSON for knowledge serialization. A measurement restrict was additionally imposed to mitigate towards denial-of-service assaults.

ChatGPT and Reminiscence Corruption: Binary File Parsing

In one other experiment, we tasked ChatGPT with producing a parser for the GGUF binary format, widely known as difficult to parse securely. GGUF recordsdata retailer mannequin weights for modules applied in C and C++, and we particularly selected this format as Databricks has beforehand discovered a number of vulnerabilities within the official GGUF library.

ChatGPT shortly produced a working implementation that accurately dealt with file parsing and metadata extraction, which is proven within the supply code under.

Nonetheless, upon nearer examination, we found vital safety flaws associated to unsafe reminiscence dealing with. The generated C/C++ code included unchecked buffer reads and cases of kind confusion, each of which may result in reminiscence corruption vulnerabilities if exploited.

On this GGUF parser, a number of reminiscence corruption vulnerabilities exist as a result of unchecked enter and unsafe pointer arithmetic. The first points included:

1. Inadequate bounds checking when studying integers or strings from the GGUF file. These may result in buffer overreads or buffer overflows if the file was truncated or maliciously crafted.
2. Unsafe reminiscence allocation, resembling allocating reminiscence for a metadata key utilizing an unvalidated key size with 1 added to it. This size calculation can integer overflow leading to a heap overflow.

An attacker may exploit the second of those points by crafting a GGUF file with a pretend header, an especially giant or destructive size for a key or worth subject, and arbitrary payload knowledge. For instance, a key size of 0xFFFFFFFFFFFFFFFF (the utmost unsigned 64-bit worth) may trigger an unchecked malloc() to return a small buffer, however the subsequent memcpy() would nonetheless write previous it leading to a basic heap primarily based buffer overflow. Equally, if the parser assumes a legitimate string or array size and reads it into reminiscence with out validating obtainable house, it may leak reminiscence contents. These flaws may doubtlessly be used to attain arbitrary code execution.

To validate this problem, we tasked ChatGPT to generate a proof-of-concept that creates a malicious GGUF file and passes it into the susceptible parser. The ensuing output reveals this system crashing contained in the memmove perform, which is executing the logic comparable to the unsafe memcpy name. The crash happens when this system reaches the top of a mapped reminiscence web page and makes an attempt to put in writing past it into an unmapped web page, triggering a segmentation fault as a result of an out-of-bounds reminiscence entry.

As soon as once more we adopted up by asking ChatGPT for solutions on fixing the code and it was capable of recommend the next enhancements:

We then took the up to date code and handed the proof of idea GGUF file to it and the code detected the malformed report.

Once more, the core problem wasn’t ChatGPT’s capacity to generate purposeful code, however moderately that the informal method inherent to vibe coding allowed harmful assumptions to go unnoticed within the generated implementation.

Prompting as a Safety Mitigation

Whereas there isn’t any substitute for a safety skilled reviewing your code to make sure it is not susceptible, a number of sensible, low-effort methods can assist mitigate dangers throughout a vibe coding session. On this part, we describe three simple strategies that may considerably cut back the probability of producing insecure code. Every of the prompts introduced on this submit was generated utilizing ChatGPT, demonstrating that any vibe coder can simply create efficient security-oriented prompts with out in depth safety experience.

Normal Safety-Oriented System Prompts

The primary method includes utilizing a generic, security-focused system immediate to encourage the LLM towards safe coding behaviors from the outset. Such prompts present baseline safety steerage, doubtlessly bettering the protection of the generated code. In our experiments, we utilized the next immediate:

Language or Software-Particular Prompts

When the programming language or utility context is understood upfront, one other efficient technique is to supply the LLM with a tailor-made, language-specific or application-specific safety immediate. This methodology immediately targets recognized vulnerabilities or widespread pitfalls related to the duty at hand. Notably, it isn’t even mandatory to concentrate on these vulnerability courses explicitly, as an LLM itself can generate appropriate system prompts. In our experiments, we instructed ChatGPT to generate language-specific prompts utilizing the next request:

Self-Reflection for Safety Assessment

The third methodology incorporates a self-reflective evaluation step instantly after code technology. Initially, no particular system immediate is used, however as soon as the LLM produces a code part, the output is fed again into the mannequin to explicitly establish and deal with safety vulnerabilities. This method leverages the mannequin’s inherent capabilities to detect and proper safety points which will have been initially missed. In our experiments, we supplied the unique code output as a person immediate and guided the safety evaluation course of utilizing the next system immediate:

Empirical Outcomes: Evaluating Mannequin Conduct on Safety Duties

To quantitatively consider the effectiveness of every prompting method, we performed experiments utilizing the Safe Coding Benchmark from PurpleLlama’s Cybersecurity Benchmark’s testing suite. This benchmark contains two varieties of exams designed to measure an LLM’s tendency to generate insecure code in eventualities immediately related to vibe coding workflows:

• Instruct Assessments: Fashions generate code primarily based on express directions.
• Autocomplete Assessments: Fashions predict subsequent code given a previous context.

Testing each eventualities is especially helpful since, throughout a typical vibe coding session, builders typically first instruct the mannequin to supply code after which subsequently paste this code again into the mannequin to deal with points, intently mirroring instruct and autocomplete eventualities respectively. We evaluated two fashions, Claude 3.7 Sonnet and GPT 4o, throughout all programming languages included within the Safe Coding Benchmark. The next plots illustrate the share change in susceptible code technology charges for every of the three prompting methods in comparison with the baseline situation with no system immediate. Unfavorable values point out an enchancment, that means the prompting technique lowered the speed of insecure code technology.

Claude 3.7 Sonnet Outcomes

When producing code with Claude 3.7 Sonnet, all three prompting methods supplied enhancements, though their effectiveness different considerably:

• Self Reflection was the simplest technique general. It lowered insecure code technology charges by a mean of 48% within the instruct situation and 50% within the autocomplete situation. In widespread programming languages resembling Java, Python, and C++, this technique notably lowered vulnerability charges by roughly 60% to 80%.
• Language-Particular System Prompts additionally resulted in significant enhancements, lowering insecure code technology by 37% and 24%, on common, within the two analysis settings. In almost all circumstances, these prompts had been more practical than the generic safety system immediate.
• Generic Safety System Prompts supplied modest enhancements of 16% and eight%, on common. Nonetheless, given the larger effectiveness of the opposite two approaches, this methodology would typically not be the beneficial selection.

Though the Self Reflection technique yielded the most important reductions in vulnerabilities, it may possibly generally be difficult to have an LLM evaluation every particular person part it generates. In such circumstances, leveraging Language-Particular System Prompts could supply a extra sensible different.

GPT 4o Outcomes

• Self Reflection was once more the simplest technique general, lowering insecure code technology by a mean of 30% within the instruct situation and 51% within the autocomplete situation.
• Language-Particular System Prompts had been additionally extremely efficient, lowering insecure code technology by roughly 24%, on common, throughout each eventualities. Notably, this technique sometimes outperformed self reflection within the instruct exams with GPT 4o.
• Generic Safety System Prompts carried out higher with GPT 4o than with Claude 3.7 Sonnet, lowering insecure code technology by a mean of 13% and 19% within the instruct and autocomplete eventualities respectively.

Total, these outcomes clearly show that focused prompting is a sensible and efficient method for bettering safety outcomes when producing code with LLMs. Though prompting alone isn’t a whole safety answer, it offers significant reductions in code vulnerabilities and might simply be personalized or expanded in keeping with particular use circumstances.

Impression of Safety Methods on Code Technology

To higher perceive the sensible trade-offs of making use of these security-focused prompting methods, we evaluated their impression on the LLMs’ normal code-generation skills. For this objective, we utilized the HumanEval benchmark, a widely known analysis framework designed to evaluate an LLM’s functionality to supply purposeful Python code within the autocomplete context.

The desk above reveals the share change in HumanEval success charges for every safety prompting technique in comparison with the baseline (no system immediate). For Claude 3.7 Sonnet, all three mitigations both matched or barely improved baseline efficiency. For GPT 4o, safety prompts reasonably decreased efficiency, aside from the Python-specific immediate, which matched baseline outcomes. Nonetheless, given these comparatively small variations in comparison with the substantial discount in susceptible code technology, adopting these prompting methods stays sensible and helpful.

The Rise of Agentic Coding Assistants

A rising variety of builders are transferring past conventional IDEs and into new, AI-powered environments that provide deeply built-in agentic help. Instruments like Cursor, Cline, and Claude-Code are a part of this rising wave. They transcend autocomplete by integrating linters, take a look at runners, documentation parsers, and even runtime evaluation instruments, all orchestrated via LLMs that act extra like brokers than static copilot fashions.

These assistants are designed to motive about your whole codebase, make clever solutions, and repair errors in actual time. In precept, this interconnected toolchain ought to enhance code correctness and safety. In follow, nonetheless, our pink workforce testing reveals that safety vulnerabilities nonetheless persist, particularly when these assistants generate or refactor advanced logic, deal with enter/output routines, or interface with exterior APIs.

We evaluated Cursor in a security-focused take a look at just like our earlier evaluation. Ranging from scratch, we prompted Claude 4 Sonnet with: “Write me a primary parser for the GGUF format in C, with the power to load or write a file from reminiscence.” Cursor autonomously browsed the net to collect particulars in regards to the format, then generated a whole library that dealt with GGUF file I/O as requested. The outcome was considerably extra sturdy and complete than code produced with out the agentic circulation. Nonetheless, throughout a evaluation of the code’s safety posture, a number of vulnerabilities had been recognized, together with the one current within the read_str() perform proven under.

Right here, the str->n attribute is populated immediately from the GGUF buffer and used, with out validation, to allocate a heap buffer. An attacker may provide a maximum-size worth for this subject which, when incremented by one, wraps round to zero as a result of integer overflow. This causes malloc() to succeed, returning a minimal allocation (relying on the allocator’s habits), which is then overrun by the following memcpy() operation, resulting in a basic heap-based buffer overflow.

Mitigations

Importantly, the identical mitigations we explored earlier on this submit: security-focused prompting, self-reflection loops, and application-specific steerage, proved efficient at lowering susceptible code technology even in these environments. Whether or not you are vibe coding in a standalone mannequin or utilizing a full agentic IDE, intentional prompting and post-generation evaluation stay mandatory for securing the output.

Self Reflection

Testing self-reflection inside the Cursor IDE was simple: we merely pasted our earlier self-reflection immediate immediately into the chat window.

This triggered the agent to course of the code tree and seek for vulnerabilities earlier than iterating and remediating the recognized vulnerabilities. The diff under reveals the result of this course of in relation to the vulnerability we mentioned earlier.

Leveraging .cursorrules for Safe-By-Default Technology

Certainly one of Cursor’s extra highly effective however lesser-known options is its assist for a .cursorrules file inside the supply tree. This configuration file permits builders to outline customized steerage or behavioral constraints for the coding assistant, together with language-specific prompts that affect how code is generated or refactored.

To check the impression of this function on safety outcomes, we created a .cursorrules file containing a C-specific safe coding immediate, as per our earlier work above. This immediate emphasised secure reminiscence dealing with, bounds checking, and validation of untrusted enter.

After putting the file within the root of the venture and prompting Cursor to regenerate the GGUF parser from scratch, we discovered that most of the vulnerabilities current within the authentic model had been proactively prevented. Particularly, beforehand unchecked values like str->n had been now validated earlier than use, buffer allocations had been size-checked, and using unsafe capabilities was changed with safer options.

For comparability, right here is the perform that was generated to learn string varieties from the file.

This experiment highlights an essential level: by codifying safe coding expectations immediately into the event surroundings, instruments like Cursor can generate safer code by default, lowering the necessity for reactive evaluation. It additionally reinforces the broader lesson of this submit that intentional prompting and structured guardrails are efficient mitigations even in additional refined agentic workflows.

Apparently, nonetheless, when working the self-reflection take a look at described above on the code tree generated on this method, Cursor was nonetheless capable of detect and remediate some susceptible code that had been missed throughout technology.

Integration of Safety Instruments (semgrep-mcp)

Many agentic coding environments now assist the combination of exterior instruments to reinforce the event and evaluation course of. One of the crucial versatile strategies for doing that is via the Mannequin Context Protocol (MCP), an open commonplace launched by Anthropic that allows LLMs to interface with structured instruments and providers throughout a coding session.

To discover this, we ran an area occasion of the Semgrep MCP server and linked it on to Cursor. This integration allowed the LLM to invoke static evaluation checks on newly generated code in actual time, surfacing safety points resembling using unsafe capabilities, unchecked enter, and insecure deserialization patterns.

To perform this, we ran the server domestically with the command: `uv run mcp run server.py -t sse` after which added the next json to the file ~/.cursor/mcp.json:

Lastly, we created a .customrules file inside the venture containing the immediate: “Carry out a safety scan of all generated code utilizing the semgrep device”. After this we used the unique immediate for producing the GGUF library, and as might be seen within the screenshot under, Cursor robotically invokes the device when wanted.

The outcomes had been encouraging. Semgrep efficiently flagged a number of of the vulnerabilities in earlier iterations of our GGUF parser. Nonetheless, what stood out was that even after the semgrep automated evaluation, making use of self-reflection prompting nonetheless uncovered further points that had not been flagged by static evaluation alone. These included edge circumstances involving integer overflows and refined misuses of pointer arithmetic, that are bugs that required deeper semantic understanding of the code and context.

This dual-layer method, combining automated scanning with structured LLM-based reflection, proved particularly highly effective. It highlights that whereas built-in instruments like Semgrep elevate the baseline for safety throughout code technology, agentic prompting methods stay important for catching the complete spectrum of vulnerabilities, particularly people who contain logic, state assumptions, or nuanced reminiscence habits.

Conclusion: Vibes Aren’t Sufficient

Vibe coding is interesting. It is quick, pleasant, and sometimes surprisingly efficient. Nonetheless, in terms of safety, relying solely on instinct or informal prompting is not adequate. As we transfer towards a future the place AI-driven coding turns into commonplace, builders should study to immediate with intention, particularly when constructing programs which are networked, unmanaged code, or extremely privileged code.

At Databricks, we’re optimistic in regards to the energy of generative AI – however we’re additionally life like in regards to the dangers. By way of code evaluation, testing, and safe immediate engineering, we’re constructing processes that make vibe coding safer for our groups and our prospects. We encourage the business to undertake comparable practices to make sure that pace doesn’t come at the price of safety.

To study extra about different finest practices from the Databricks Purple Crew, see our blogs on find out how to securely deploy third-party AI fashions and GGML GGUF File Format Vulnerabilities.