claw-code/.guardrails/docs/agentmcp/Project Readme.txt

27 lines
2.8 KiB
Plaintext
Raw Blame History

This file contains invisible Unicode characters

This file contains invisible Unicode characters that are indistinguishable to humans but may be processed differently by a computer. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.

Project Sentinel: The Active Guardrail PlatformVersion: 3.0.0-EnterpriseStatus: Production ReadyArchitecture: Sentinel Core🛡 What Is Project Sentinel?Project Sentinel is an active governance layer for Autonomous AI Agents. Unlike passive templates that rely on an LLM "remembering" to follow rules, Sentinel uses a compiled Model Context Protocol (MCP) server to physically enforce safety, security, and financial guardrails.It transforms the "Soft Laws" of documentation into the "Hard Physics" of the development environment.Core CapabilitiesVFS Jail: Prevents Agents from accessing restricted paths (.env, ~/.ssh).State Machine: Enforces strict SDLC states (Planning -> Active -> Review).FinOps: Tracks token usage and enforces budget caps ($/sprint).Polyglot: Native toolchains for Go, Python, TS, Rust, Java, and 10+ others.📂 Repository StructureThe repository is organized into specific domains. Agents should query the INDEX_MAP.md to find specific rules..
├── bin/ # Compiled Sentinel binaries
├── services/
│ └── sentinel/ # The Sentinel Core (Go Source)
│ ├── cmd/ # Entry points
│ ├── internal/ # Kernel, Jailor, Auditor, Ledger
│ └── tools/ # MCP Tool Definitions
├── docs/
│ ├── SENTINEL_ARCH.md # System Architecture
│ ├── workflows/ # Operational Protocols (Commit, Deploy, Review)
│ ├── standards/ # Coding Standards (API, Logging, Secrets)
│ ├── languages/ # Language-Specific Profiles (Go, Python, etc.)
│ └── setup/ # Installation & Config
├── sprints/ # Active Sprint Data (if not using DB)
└── sentinel.toml # The Policy Configuration
🚀 Quick Start1. Installation# Download and install the background service
curl -L [https://releases.project-sentinel.io/install.sh](https://releases.project-sentinel.io/install.sh) | sh
sentinel init --root .
2. Connect Your AgentAdd the Sentinel MCP server to your Agent configuration (Claude Desktop, OpenCode, Cursor):{
"mcpServers": {
"sentinel": {
"command": "./bin/sentinel",
"args": ["serve", "--stdio"]
}
}
}
3. Start a SprintAsk your Agent:"Initialize a new sprint for the Auth Service refactor. Create tasks for Login, Logout, and Password Reset."🤖 Agent Instructions: Maintaining This READMEProtocol for AI Agents:Trigger: When you add a new top-level directory or significantly change the architecture.Action: You must update the Repository Structure tree above.Constraint: Do not remove the INDEX_MAP link.Verification: Ensure sentinel --version matches the version badge at the top.License: BSD-3-Clause | Maintainer: TheArchitectit