Introducing deepsec: A Security Tool for Discovering Vulnerabilities in Codebases

Today we’re open sourcing deepsec: a security harness powered by coding agents. It runs on your own infrastructure and surfaces hard-to-find issues in large codebases.

You can run deepsec on your laptop without setting up a cloud service for privileged source code access. For inference, you can use your existing Claude or Codex subscription without any additional setup.

Scanning large repos can take multiple days on a single machine. To run research jobs in parallel, deepsec supports optional fanout to Vercel Sandboxes for remote execution. Scans on Vercel’s codebases routinely scale up to 1,000+ concurrent sandboxes.

Architecture

At its core, deepsec uses claude and codex to perform tailored investigation of a codebase using Opus 4.7 at max effort and GPT 5.5 at xhigh reasoning.

Scans start with static analysis to identify security-sensitive files, then coding agents investigate each candidate, tracing data flows, checking for mitigations, and producing actionable findings with severity ratings. Here is the workflow:

Scan: It starts by performing a regex-only scan of all files for security-sensitive areas that subsequent steps will focus on.
Investigate: Agents investigate each file identified in the scan.
Revalidate: A second agent run validates investigation findings to remove false positives and reclassify severity.
Enrich: Once investigation is complete, an agent uses git metadata and other optional services to identify the contributors responsible for fixing each issue.
Export: The export command formats the findings as instructions so that they can be turned into tickets for humans and coding agents.

Running `deepsec` on production code

deepsec has been highly useful on our own monorepos and for our customers' codebases. During development, we ran deepsec on several open source repositories of Vercel customers and partners.

For example, deepsec scanned the open source version of dub.co. Dub is a marketing attribution platform for affiliate programs and short links that is also available as SaaS. It features authenticated access, interacts with a database, and runs several backend services, creating a large security surface. When we shared our deepsec findings with founder Steven Tey, he replied:

Running against Vercel’s own monorepos, deepsec identified subtle edge cases in auth conditions, leading us to develop a custom scanner plugin that covers every authentication path in our code.

False positives and best uses

Some of deepsec's findings will be false positives. In our experience the false positive rate is roughly 10-20%. Given the impact of true positive findings in our own research, we’ve been happy with this outcome, and we built the revalidate step to have the agent further verify its findings to reduce false positives.

deepsec works best for applications and services. It may be usable for libraries and frameworks, but those would likely require custom prompts and scanners.

Customization and plugins

deepsec ships with a plugin system for adapting it to your codebase. The most common plugins are custom scanners: regex matchers tuned to your auth model, data layer, or team conventions. We recommend using deepsec with your coding agent and asking it to write those matchers based on findings from an initial scan:

Do I need access to a special “cyber model”?

Both Anthropic and OpenAI offer “cyber” versions of their most capable models, fine-tuned to accept security tasks the base models won’t. deepsec works with these, but is also fully functional with off-the-shelf models.

deepsec ships with a classifier that checks whether the task was refused after each research step. In our experience, for the prompt that deepsec is using, refusals are a non-issue for both Opus 4.7 and GPT 5.5.

Getting started

To get started, run npx deepsec init at the root of your repository. This will create a directory called ./.deepsec, which is used to configure the system and store a catalog of your deepsec investigations. From there, follow the output of the command. Read the full documentation on Github.

Feedback welcome

While we’ve used deepsec extensively, it is still early in its development. Feedback and contributions on GitHub are welcome.