Keep your AI coding standards in one place.
Install them as native config everywhere.
Everything braid uses to turn standards into a delivery system. Author rules and skills once, then run braid install to compile native config for Cursor, Claude Code, Copilot, Windsurf, and more. Share them across your team so every engineer's AI tools enforce the same standards.
Standards Authoring
The braid editor gives you a rich authoring environment purpose-built for coding standards. Define rules that encode how your team writes code, or create skills that sub-agents invoke on demand. Author once in braid, then deliver as native config to Cursor, Claude Code, Copilot, Windsurf, and 15+ other tools — no manual copying between repos or format conversion.
- Rich editor with AI-assisted drafting
- Author once, deliver to 15+ tool formats
- Version history and change tracking
Cross-Tool Delivery
braid compiles your rules and skills into the native config format each tool expects — .cursor/rules/ for Cursor, .claude/rules/ and .claude/skills/ for Claude Code, .github/copilot-instructions.md for Copilot, and more. Run braid install or connect via MCP, and every engineer's toolchain picks up the latest standards automatically. No more maintaining parallel config files across repos or hoping everyone copies the right version.
- Compiles to native config for 15+ AI tools
- CLI install and MCP live-sync delivery
- Eliminates repo-per-tool config sprawl
Shared Libraries
Share standards and sub-agent configurations across your entire team from a single library. When one engineer updates a rule, every team member's AI tools get the update automatically — no pull requests or manual syncs required. Role-based permissions let you control who can author, review, and publish standards, so your library stays curated.
- Shared organization standards library
- One update reaches every team member
- Role-based access control
Version History
braid keeps a full version history for every rule and skill in your library. Compare any two versions side-by-side, see exactly who made each change, and restore a previous version instantly if a standard regresses. Unlike git-only approaches, version history is built into the authoring flow — no separate commits or PRs needed to track what changed.
- Full edit history for every standard
- Side-by-side version comparison
- One-click rollback to any version
Specialist Sub-agents
Define specialist sub-agents for different types of work — a frontend sub-agent, a backend sub-agent, a testing sub-agent — each with their own model preferences, system prompts, and skill references. When a sub-agent starts a conversation, it automatically loads the right context for its specialty.
- Role-specific sub-agents with custom context
- Per-sub-agent model and prompt configuration
- Shared skills across sub-agent types
Multi-Step Workflows
Workflows let you standardize repeatable multi-step AI tasks. Define a sequence of prompts, conditions, and checkpoints, then publish it for your team. Anyone can run the workflow and get consistent outcomes — whether it's a release checklist, code review process, or onboarding procedure.
- Visual multi-step execution graphs
- Checkpoints and resume support
- Consistent outcomes across team members
Project Scoping
Group your prompts by project. Control exactly which prompts are active for each project, making it easy to manage standards across multiple codebases. Sub-agents connected to a project only see the prompts that are relevant to their work.
- Project-scoped prompt libraries
- Active/inactive toggle per prompt
- Multi-project support
Automated Triggers
Set up webhooks, cron schedules, and workflow chains to automate your AI workflows. Trigger workflows from external events, run them on a schedule, or chain them together so one workflow kicks off the next. Build complex automation pipelines without writing glue code.
- Webhook and cron-based automation
- Workflow chaining and cascading triggers
- Event-driven execution pipelines
Built-in AI Chat
Use braid's built-in chat to research patterns, draft new prompts and refine existing ones using top AI models. Access multiple models through OpenRouter, compare outputs side-by-side, and iterate on your prompts until they're production-ready.
- Access multiple models via OpenRouter
- Draft and refine prompts with AI assistance
- Research patterns and best practices