Skip to main content
bex.co
419

Kimi Code security scan

Run a Kimi Code security scan on your codebase.

Use Kimi Code's native ACP agent—or Kimi behind Claude Code—while Bex Security validates findings and preserves reviewable scan artifacts.

Install and run a scanExplore Bex Security

Open source · Apache 2.0 · Report-only by default

Kimi Code security scan
npm install --global @bex-co/bex-security
kimi login
bex-security scan . --agent kimi --model kimi-code/k3-256k --effort high
Kimi CodeKimi K3Bex Security

Kimi K3

Selected model

Kimi Code

Agent runtime

Bex Security

Security orchestration

More than a prompt

What a Kimi security scan actually reviews.

Kimi investigates the repository; Bex supplies the repeatable workflow that turns model reasoning into evidence an engineer can verify.

  1. 01

    Scope the repository

    Define source boundaries and permissions before investigation begins.

  2. 02

    Discover attack paths

    Use Kimi K3 to trace likely vulnerability paths across the scoped codebase.

  3. 03

    Validate evidence

    Challenge candidate findings and reject claims that lack support.

  4. 04

    Seal scan artifacts

    Preserve reviewable findings, evidence, and coverage as local artifacts.

The Bex layer

Use Kimi without weakening the security review.

Choose native Kimi Code or the Claude provider path without changing what counts as a complete, validated scan.

Open agent boundary

Native Kimi speaks ACP directly; the alternate provider path runs Kimi behind Claude Code.

Stable scan contract

Bex keeps repository scope, workbench tools, validation, coverage, and final artifacts consistent.

Scoped credentials

Native scans use Kimi's saved login; the Claude provider path receives KIMI_API_KEY only for the current process.

Quick setup

How to run a native Kimi Code security scan.

Install Bex Security and Kimi Code, log in once, then select the native Kimi agent and an available model.

Prefer Claude Code as the agent shell?

Use the Kimi provider path with a scoped API key; it defaults to kimi-for-coding.

export KIMI_API_KEY="<your-kimi-api-key>"
bex-security scan . --agent claude --provider kimi
1

Install both CLIs

Install @bex-co/bex-security from npm and follow Kimi's official installation guide so kimi is on PATH.

2

Log in to Kimi Code

Run kimi login once. Native scans use Kimi Code's saved authentication and provider settings.

3

Select the Kimi agent

Pass --agent kimi and optionally choose the model and reasoning effort advertised by the Kimi ACP session.

Alpha integration. Native Kimi ACP and Kimi through Claude Code are available for early use while full-scan compatibility coverage continues to expand.

Frequently asked questions

Should I use native Kimi Code or Kimi through Claude Code?

Use --agent kimi when you want Kimi Code to own authentication, model selection, and the ACP session. Use --agent claude --provider kimi when your team wants to keep Claude Code as the agent shell.

Which credentials does a Kimi security scan use?

Native scans use the login saved by Kimi Code. The Claude provider route reads KIMI_API_KEY for the current scan and Bex does not store that key.

Will Kimi change my repository during a security scan?

Not during an ordinary scan. Bex keeps scans report-only unless you explicitly request patching after findings have been validated.

Explore other AI security scans

Make Kimi code security reviews repeatable.

Install Bex Security when you want Kimi's code reasoning inside a workflow that validates findings and preserves evidence.

Star Bex Security

One workflow · Open agent layer · Validated evidence