{
  "id": 7867456,
  "title": "I tried giving a 26 MB Burp Suite dump to an LLM. It didn't work. So I built BurpSqueezer.",
  "url": "https://urgent.news/2026/09/16/i-tried-giving-a-26-mb-burp-suite-dump-to-an-llm-it-didnt-work-so-i",
  "topic": "ai",
  "section": "AI",
  "published": "2026-09-16T20:27:54.000Z",
  "source": {
    "name": "Dev.to",
    "slug": "dev-to",
    "url": "https://dev.to/vaginskii/i-tried-giving-a-26-mb-burp-suite-dump-to-an-llm-it-didnt-work-so-i-built-burpsqueezer-29i4"
  },
  "original_language": "en",
  "account": "I recently tested the capability of a large language model (LLM) with a 26 MB dump created by Burp Suite. Unfortunately, the LLM struggled to process the large file size, which led to the creation of BurpSqueezer. This new tool aims to help web and API penetration testers integrate AI into their workflow.\n\nBurpSqueezer compresses HTTP traffic, which can range from tens to hundreds of megabytes, into a small .md report specifically designed for AI processing. This compressed report is incredibly small, with a test case reducing a 26.7 MB dump to just 35 KB—745 times smaller than the original. The smaller size allows almost any neural network to read and understand the report.\n\nHowever, there are some limitations to BurpSqueezer. It is primarily designed for large HTTP datasets, particularly those involved in APIs and applications with meaningful business logic. Smaller or structurally simple dumps, such as websites with minimal backend logic and mostly static or repetitive traffic, may not yield as useful results due to a lack of relationships and structure for analysis.\n\nBurpSqueezer is not an autonomous pentesting tool. It does not send requests or attack the target; instead, it takes pre-captured and authorized traffic, processes it, and produces a more compact representation that can then be given to an LLM for further analysis. The project is open source, and the author, Vaginskii, welcomes feedback from those working with Burp, API security, business logic, or AI-assisted pentesting. The GitHub repository for BurpSqueezer is located at vaginskii/BurpSqueezer.",
  "summary": "Hi, everyone. I’ve put together a tool that will help many web and API penetration testers solve the problem of integrating AI into their workflow. BurpSqueezer is a utility that lets you “compress” the HTTP traffic you need for your work (which can take up tens or even hundreds of megabytes) into a small .md report, generated specifically for AI processing. I haven’t hardcoded anything into the…",
  "key_points": [],
  "editors_take": null,
  "illustration": null,
  "coverage": {
    "outlets": 1,
    "also_reported_by": []
  },
  "ai_generated": true,
  "disclaimer": "Summaries, key points and the editor’s take are written by software from other outlets’ reporting and may contain errors — always check the linked original."
}