Free EARY ACCESS TRIAL
Start governing your AI-generated code today. Free.
Get Salt Code with four ready-to-use security policy packs. Forty policies. Every AI coding assistant your developers use. From the first prompt to your first compliant commit, no cost to start.
- Four secure coding packs included
- 40+ policies enforced day one
- Works with Cursor, Copilot, Claude, and every MCP-supported assistant
- Deploys in minutes — no credit card required
Already a Salt Security customer? Salt Code is included in your existing license at no additional cost. Contact your account team to activate.
Request free access to Salt Code
4
Secure coding packs included
40+
Policies enforced
day one
10+
AI coding assistants supported
$0
To get started
How it works Three steps. Compliant code by default.
No new tools to install. No developer workflow change. Connect Salt Code via MCP, pick your packs, and your AI coding assistants start generating policy-compliant code.
01 Connect
Add Salt Code as an MCP server in Cursor, GitHub Copilot, Claude, or any assistant your developers use. One configuration, every assistant.
02 Activate packs
Turn on the four packs that match your needs. OWASP API. MCP Security. LLM Security. OpenAPI compliance. Mix and match. Day one coverage.
03 Generate
Developers prompt their AI assistants the same way they always do. The output arrives policy-compliant. No backlog, no PR rewrites, no surprise findings later.
What you get Four secure coding packs. Forty policies. Day one.
Every Salt Code free account ships with four pre-built security policy packs, each grounded in industry-standard or Salt-defined frameworks. No policy authoring required. Drop them in, connect your AI coding assistant, and your generated code is compliant from the first prompt.
Pack 01 · API Security OWASP API Top 10
The industry-standard playbook for API security risks, translated into enforceable rules for every API your AI assistants generate. Catches broken authorization, unrestricted resource consumption, and the rest of the top ten at the point of code creation.
- Broken object level authorization checks enforced
- Unrestricted resource consumption protections in place
- Server-side request forgery prevention applied
Pack 02 · Agentic Security MCP Security Top 10
Salt-defined security policies for MCP server implementations, agent tool registration, and the unique attack patterns of agentic infrastructure. The only ready-to-use ruleset for the MCP layer your agents now depend on.
- MCP server authentication and scope enforcement
- Tool description validation against prompt injection
- Least-privilege scope on registered agent tools
Pack 03 · LLM Security LLM Security Top 10
Based on the OWASP Top 10 for LLM applications. Catches prompt injection patterns, sensitive information disclosure, excessive agency, and the rest of the LLM-specific risks at code generation time, before the AI assistant even commits the code.
- Prompt injection prevention on user-facing inputs
- Sensitive information disclosure controls
- Excessive agency limitations on agent tool use
Pack 04 · Spec Compliance OpenAPI / Swagger Compliance
Every AI-generated API conforms to the OpenAPI specification with documented authentication schemes, response codes, and schema definitions. Stop shipping undocumented endpoints and ad-hoc spec sprawl into production.
- Authentication scheme documented on every endpoint
- Required and optional fields explicitly annotated
- Response codes specified for all paths
What’s included Free access, production-ready.
This is not a trial that times out. This is real Salt Code, with real policy packs, running against real AI coding assistants. For non-customers, the path to compliant AI-generated code starts at zero cost.
All four secure coding packs
Forty policies enforced from day one across every AI coding assistant in your stack.
Every AI coding assistant
Cursor, GitHub Copilot, Claude, Codex, Gemini CLI, Windsurf, Kiro, Antigravity, and every other MCP-supported tool.
Real-time policy enforcement
Code generated inside your developers' AI assistants arrives compliant. No after-the-fact review needed.
The Salt Posture Governance Engine
The same engine that powers the Salt Agentic Security Platform for our enterprise customers.
No credit card to start
Sign up, connect, generate. Upgrade to custom policies and platform features when you are ready.
Path to the full platform
Start with the packs. Expand into CI/CD validation, runtime monitoring, and the full Salt Agentic Security Platform when the time is right.
Every line of AI-generated code is a policy decision.
Start making yours the right one. Salt Code is available today.
Early Access Program:
Free access is open to the first 100 organizations to enroll. All four pre-built Secure Coding Packs included — OWASP API Top 10, MCP Security Top 10, LLM Security Top 10, and OpenAPI/Swagger Compliance.