windsurf-security-basics
Instructions for excluding sensitive files from AI indexing and maintaining a secure local development environment.
Install
mkdir -p .claude/skills/windsurf-security-basics && curl -L -o skill.zip "https://agentskills.codes/api/skills/download/4834" && unzip -o skill.zip -d .claude/skills/windsurf-security-basics && rm skill.zipInstalls to .claude/skills/windsurf-security-basics
Activation
This is the description your AI agent reads to decide when to run this skill — the better it matches your request, the more reliably it fires.
Apply Windsurf security best practices for workspace isolation, dataKey capabilities
- →Exclude sensitive files from AI indexing
- →Disable telemetry for privacy compliance
- →Configure AI autocomplete exclusions for secret-containing file types
- →Create security-focused .windsurfrules
- →Audit AI workspace access for potential secret exposure
How it works
The skill uses `.codeiumignore` to exclude files, modifies Windsurf settings to disable telemetry and autocomplete for certain file types, and suggests security rules in `.windsurfrules`.
Inputs & outputs
When to use windsurf-security-basics
- →Protect credentials from AI leakage
- →Configure .codeiumignore for projects
- →Audit AI workspace access
- →Implement least-privilege for AI tools
About this skill
Devin Desktop Security Basics
Overview
Devin Desktop is the current name for Windsurf. Secure it with layered controls: repository permissions and secret management are authoritative; .codeiumignore, Rules, and prompts reduce context exposure but do not replace access control.
Prerequisites
- A version-controlled workspace and clean working tree
- The organization's data-classification and secret-handling policy
- Authorization to inspect configuration without collecting secret values
Tool Use
- Use
Readto inspect only the repository files and configuration needed for the request. - Use
Grepto locate relevant settings, rules, logs, or code without broad collection. - Use
Writeonly for a new artifact the user requested; never write credentials or unreviewed production configuration.
Instructions
Step 1: Map trust boundaries
Record repository sensitivity, connected organizations, enabled models, MCP servers, Hooks, terminal permissions, remote indexing, and deployment integrations. Identify which controls are local, repository-owned, or centrally administered.
Step 2: Configure context exclusions
Create .codeiumignore at the repository root using gitignore syntax. Include only project-relevant patterns, for example:
.env
.env.*
*.pem
*.key
credentials/
secrets/
customer-data/
dist/
node_modules/
Windsurf also honors .gitignore, and Enterprise administrators can add a global .codeiumignore under ~/.codeium/. Verify behavior with a harmless canary filename; do not place a real secret in the test.
Step 3: Enforce durable behavior
Use repository AGENTS.md or .devin/rules/*.md to require secure coding, validation, least privilege, and explicit approval for risky operations. Keep each workspace Rule within the documented 12,000-character limit and use the correct trigger: mode.
Step 4: Review tools and automation
- Disable unused MCP servers and tools.
- Prefer provider OAuth or environment-backed secrets over literal values in
mcp_config.json. - Review Hooks as executable policy code and fail closed where required.
- Keep protected branches, code owners, CI, and deployment approvals outside the model's discretion.
- Require human approval for production, destructive, identity, and billing mutations.
Step 5: Respond to exposure
If a secret appears in a prompt, output, log, diff, or diagnostic bundle, treat it as exposed: stop sharing, revoke or rotate it through the provider, remove it from history where authorized, and document the incident. Adding an ignore rule alone is not remediation.
Step 6: Verify
Confirm ignored paths are absent from context, protected branches still require review, MCP tools match policy, Hooks run on representative success and failure paths, and no configuration file contains credential material.
Output
Deliver a security review with context boundaries, ignored paths, organization policy, MCP and Hook exposure, repository protections, detected secret locations without values, remediation owners, and verification evidence. Mark controls as preventive, detective, or corrective.
Error Handling
| Issue | Response |
|---|---|
| Ignored file still appears | Check syntax, location, gitignore interaction, and current indexing state |
| MCP requires a token | Use provider OAuth or an approved secret store; never commit the value |
| Hook blocks legitimate work | Preserve the event, fix the narrow rule, and retest failure handling |
| Compliance claim is uncertain | Link first-party evidence and route the decision to security or legal |
Examples
Finding: ".env.production was not excluded. Add .env.*, rotate any exposed credential, verify with a harmless canary, and retain protected-branch review as the enforcement boundary."
Resources
Related Skill
Continue with windsurf-data-handling to build a regulated-data inventory, verify mutable vendor claims, and record control evidence without exposing sensitive values.
Prerequisites
Limitations
- →Secrets appearing in AI output if source file is not in `.codeiumignore`
- →AI indexing .env files if `.env*` pattern is not added to `.codeiumignore`
- →Telemetry sending code if all telemetry settings are not disabled
How it compares
This skill provides a structured approach to securing Windsurf AI environments against data leakage and privacy violations, which differs from a default setup that might expose sensitive information to AI indexing or telemetry.
Compared to similar skills
windsurf-security-basics side by side with the closest alternatives in the catalog.
| Skill | Installs | Updated | Safety | Difficulty |
|---|---|---|---|---|
| windsurf-security-basics (this skill) | 1 | 2mo | Review | Beginner |
| reverse-engineering-tools | 73 | 5mo | No flags | Advanced |
| game-hacking-techniques | 42 | 4mo | No flags | Advanced |
| solidity-security | 15 | 4mo | No flags | Intermediate |
Try saying
Example prompts that trigger this skill in your AI assistant.
More by jeremylongshore
View all by jeremylongshore →You might also like
reverse-engineering-tools
gmh5225
Guide for reverse engineering tools and techniques used in game security research. Use this skill when working with debuggers, disassemblers, memory analysis tools, binary analysis, or decompilers for game security research.
game-hacking-techniques
gmh5225
Guide for game hacking techniques and cheat development. Use this skill when researching memory manipulation, code injection, ESP/aimbot development, overlay rendering, or game exploitation methodologies.
solidity-security
wshobson
Master smart contract security best practices to prevent common vulnerabilities and implement secure Solidity patterns. Use when writing smart contracts, auditing existing contracts, or implementing security measures for blockchain applications.
1password
openclaw
Set up and use 1Password CLI (op). Use when installing the CLI, enabling desktop app integration, signing in (single or multi-account), or reading/injecting/running secrets via op.
senior-security
davila7
Comprehensive security engineering skill for application security, penetration testing, security architecture, and compliance auditing. Includes security assessment tools, threat modeling, crypto implementation, and security automation. Use when designing security architecture, conducting penetration tests, implementing cryptography, or performing security audits.
ghidra
mitsuhiko
Reverse engineer binaries using Ghidra's headless analyzer. Decompile executables, extract functions, strings, symbols, and analyze call graphs without GUI.