Your AI coding standards, skills, agents, and workflows distributed to every tool automatically.
Currently in private beta · Request access to join
Connect your agent to braid once. Rules inject automatically at the start of every conversation. Skills surface on demand when the context matches. Workflows help teams standardize repeatable multi-step tasks. No slash commands. Your agent just knows.
Write rules and skills in a rich editor. Import from GitHub. Organize by project. Share across your team. Define specialist agents and reusable workflows where they add leverage.
Draft and refine rules and skills with AI built into the editor. Write once, install to any tool.
Use CLI for offline access and control. Use MCP for rules that auto-inject and skills your agent invokes on demand. Or use both. Add workflows when your team needs repeatable multi-step execution.
Install a rule or skill to Cursor, Claude Code, OpenCode, and 15+ other formats with a single command. Works offline.
# Install rules + skills to current repo
$ braid install
# Update your installed prompts
$ braid updateSupports Claude Code, Cursor, OpenCode, Windsurf, Cline, Zed, and 10+ more.
Rules load at the start of every conversation. When your agent starts building, braid matches the right skills and injects them mid-conversation.
// Agent connects; rules load automatically
braid.init() // rules injected
// You ask your agent to build something
braid.get_skills("stripe checkout") // skill injectedWorks with any MCP client: Claude Code, Cursor, Claude Desktop, and more.
Rules load at conversation start. As you work, braid recognizes what you're building and surfaces the right skills. You never memorize a name or invoke anything.
Rules inject at conversation start. Skills activate on demand.
Add agents for focused roles while keeping the same shared skills underneath.
Standardize repeatable multi-step tasks with workflow runs your team can reuse.
Ask anything. Watch your skills kick in automatically.
Press Enter to send, Shift+Enter for new line
Switch between Cursor, Claude Code, or any tool. Your rules and skills follow you. One library, every format, no copying files between repos.
One engineer writes a rule or skill. Everyone's agent uses it automatically. Shared org library, project-based access, role permissions.
AI rules contain your engineering standards, security patterns, and institutional knowledge. braid gives you the controls to keep them private, auditable, and under your team's control.
Your rules never leave your control. We don't train on your data, period.
All rules and references are encrypted using AES-256. Your secrets stay secret.
Rules synced from GitHub are automatically scanned for prompt injection and malicious patterns.
Track who accessed what, when. Full visibility into rule usage across your org.
Annual billing saves 2 months. All plans include MCP and CLI access.
* Credits are based on AI usage. Cost varies by model. Cheaper models use fewer credits per message.
For individual developers. Build your own searchable rules library.
For individuals getting started
For power users
For teams. Create an organization to share rules, manage projects, and collaborate. Fully self-serve.
For small teams
For larger teams
Credits reset monthly. Top-ups available for additional capacity.
SAML and OIDC support for seamless authentication with your identity provider.
Complete visibility into rule access, changes, and usage across your organization.
Automatically sync users and groups from your identity provider via SCIM.
Role-based access control at org, project, and rule levels.
Priority support with SLA guarantees and dedicated success manager.
API access and custom MCP server configurations for your workflow.
Tell us about your team. We'll set up a 30-minute walkthrough of braid for your stack, typically within 24 hours.
Not ready yet? Stay in the loop.
Product updates only. No spam, just launches and features.
One library. Every tool. Automatic.
Request access to get started.