@pretense/cli

AI firewall CLI that mutates proprietary code identifiers before sending to LLM APIs. Your code hits AI naked -- Pretense fixes that.

Install

npm install -g @pretense/cli

Quickstart

# Initialize config in your project
pretense init

# Save a dashboard API key (read_write for mutation via proxy), or set PRETENSE_API_KEY
pretense login --key YOUR_KEY

# Start the proxy (default port 9339) — validates key before binding
pretense start

# Point your AI tools at the proxy
export ANTHROPIC_BASE_URL=http://localhost:9339
export OPENAI_BASE_URL=http://localhost:9339/v1

# Scan a file for identifiers and secrets
pretense scan src/

# Mutate a file and print to stdout
pretense mutate src/app.ts

# Reverse mutations using stored map
pretense reverse mutated-output.ts

# View audit log
pretense audit --json

# Plan / quota (live when backend reachable; falls back offline)
pretense status

Commands

| Command | Description | | --------------------------- | -------------------------------------------------------------- | | pretense init | Scan cwd, create .pretense/ config | | pretense login [--key …] | Save dashboard API key to ~/.pretense/config.json | | pretense logout | Remove saved dashboard API key | | pretense start | Validate key, start proxy on :9339 | | pretense scan <file\|dir> | Scan for identifiers + secrets | | pretense mutate <file> | Redact scanner hits (secrets/PII), mutate identifiers → stdout | | pretense reverse <file> | Reverse using stored map | | pretense audit | Print audit log | | pretense status | Plan, quota, usage (usage is an alias) | | pretense credits | Remaining mutation budget (tokens is an alias) | | pretense upgrade | Compare plans / upgrade | | pretense version | Print version |

Env: PRETENSE_API_KEY, PRETENSE_API_URL, PRETENSE_API_TIMEOUT_MS (backend HTTP timeout, default 15s).

Exit Codes

• 0 -- Success
• 1 -- Error
• 2 -- Secrets found

Learn More

Visit pretense.ai for documentation and enterprise features.