Free IDE extension for risk-free vibe coding — keep secrets out of AI.
Windsurf Security
Windsurf (by Codeium) is an agentic AI IDE that reads your project context deeply to power Cascade and other AI features. That deep context access means secrets, credentials, and personal data can silently end up in AI prompts. SoterAI IDE Guard runs locally in VS Code-compatible editors including Windsurf, scanning your workspace for secrets, PII, prompt injection, and risky MCP configurations before they reach the AI model.
Runs locally in your editor. Secret, PII, prompt-injection, and MCP scanning happen on your machine before anything reaches an AI model.
Scan your open files, selections, and workspace for API keys, tokens, database connection strings, Aadhaar patterns, PAN, and other sensitive data — all on your machine before Windsurf AI reads it.
Use 'Scan Before AI Prompt' to inspect and redact your context before sending it to Windsurf's Cascade agent or any AI command, so sensitive content never reaches the model.
Windsurf supports MCP tools via its agent mode. SoterAI IDE Guard scans MCP server configurations in your workspace and flags tools with over-broad permissions before you enable them.
Catch injected instructions in files, dependency metadata, or pasted content before they enter a Windsurf AI command — including obfuscation techniques like invisible unicode characters.
Maintain a local ledger of what Windsurf AI was given access to during a session so you can review and audit exposure without uploading anything.
Flag destructive or risky shell commands suggested by Windsurf's terminal AI before they execute — covering commands like rm -rf, curl with data upload, or environment variable exfiltration patterns.
Install SoterAI IDE Guard
Windsurf is built on VS Code's extension platform. Install SoterAI IDE Guard from the VS Code Marketplace or the Open VSX Registry directly within Windsurf.
Open your Windsurf project
The extension activates in your Windsurf workspace automatically. Local scanning is on by default — cloud features are opt-in only.
Scan before running Cascade
Before using Windsurf's Cascade agent on sensitive code, run 'Scan Before AI Prompt' from the Command Palette to surface any secrets or injection text in the context.
Redact and review
Redact detected sensitive spans in the scan results panel. Your original files are unchanged; only the context you choose to share is modified.
No security tool is perfect. Here is what this feature does not claim to do, so you can layer defenses appropriately.
Yes. Windsurf (by Codeium) is built on VS Code's extension API. SoterAI IDE Guard installs and runs in Windsurf from the VS Code Marketplace or Open VSX Registry.
Yes. Use the 'Scan Before AI Prompt' command to review your context before running Cascade. This surfaces secrets and injection text before the agent processes your request.
No. All secret, PII, and prompt-injection scanning runs locally in the extension host. Nothing leaves your machine without your explicit opt-in.
Yes. SoterAI IDE Guard's MCP scanner works with any MCP configuration file present in your workspace, including those created for Windsurf's agent tools.
Install SoterAI IDE Guard and scan secrets, prompts, MCP tools, and terminal commands locally before they ever reach an AI model.
Install for VS Code