Setup takes under 4 minutes

From Zero to Protected in 4 Minutes

Install the Pretense CLI, scan your codebase, start the local proxy, and connect your AI coding tool. Your code never leaves your machine unmutated.

Install the Pretense CLI

Install Pretense globally with npm. The proxy starts on localhost:9339 and intercepts LLM API calls before they leave your machine.

terminal

npm install -g pretense

Pretense vX.X.X installed

Pretense analyzes your project and builds a mutation profile. It detects proprietary identifiers and secrets across TypeScript, Python, Go, and Java.

terminal

pretense scan ./src

Found 47 identifiers, 2 secrets detected

Launch the Pretense proxy. It runs locally on port 9339 and intercepts all LLM API traffic from your dev tools.

terminal

pretense start

Proxy running on http://localhost:9339

Point your AI coding tool at the Pretense proxy. For Claude Code, set ANTHROPIC_BASE_URL. For OpenAI tools, set OPENAI_BASE_URL.

terminal

# For Claude Code
ANTHROPIC_BASE_URL=http://localhost:9339 claude

# For OpenAI tools
export OPENAI_BASE_URL=http://localhost:9339

First mutation complete. Your code is protected.

pretense -- live output

$ pretense start

[pretense] Proxy starting on port 9339

[pretense] Loaded mutation profile: 47 identifiers

[pretense] Intercepted: POST /v1/messages (Anthropic)

[pretense] Mutating: getUserPaymentToken -> _fn4a2b

[pretense] Mutating: stripeSecretKey -> _v8c3d

[pretense] Secret blocked: sk-ant-... (ANTHROPIC_API_KEY)

[pretense] Request forwarded. 3 identifiers mutated.

[pretense] Response received. Reversing mutation...

[pretense] Done. Round-trip: 847ms

Live mutation output. No code leaves your machine in its original form.

Free forever for individual developers. No credit card required.