RFC: Agent Mode with Security-First Design#187
Open
creative-CLAi wants to merge 3 commits intolevante-hub:developfrom
Open
RFC: Agent Mode with Security-First Design#187creative-CLAi wants to merge 3 commits intolevante-hub:developfrom
creative-CLAi wants to merge 3 commits intolevante-hub:developfrom
Conversation
Release v1.6.0-beta.2
🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
This RFC proposes adding an optional 'Agent Mode' to Levante that enables proactive, task-oriented AI capabilities while maintaining security and ease-of-use principles. Key features: - Guardian Layer: Security core that analyzes and authorizes all agent actions - Capability System: User-controlled permissions via UI toggles - MCP-based architecture: Sandboxed capabilities through existing MCP infra - Audit System: Full transparency of all agent actions Documents included: - 001-agent-mode.md: Full RFC with implementation details - 001-agent-mode-architecture.md: Visual diagrams and UI mockups Co-authored-by: CLAi <clai@clawdbot.local>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
This RFC proposes adding an optional Agent Mode to Levante that enables proactive, task-oriented AI capabilities while maintaining the security and ease-of-use principles that define Levante.
Motivation
Current AI chat interfaces are reactive. The next evolution is agentic AI — assistants that can take actions, remember context, and work proactively.
However, existing agentic tools (Clawdbot, Claude Code) are developer-focused and security-permissive by default. Levante is uniquely positioned to bring agentic AI to mainstream users (teachers, students, workers) with a security-first, user-friendly approach.
Key Components
1. Guardian Layer
Security core that intercepts and analyzes all agent actions before execution:
2. Capability System
User-controlled permissions via UI toggles:
3. MCP-Based Architecture
Leverages Levante's existing MCP infrastructure:
4. Audit System
Full transparency:
Documents
docs/rfcs/001-agent-mode.md- Full RFC with implementation details, TypeScript interfaces, Guardian promptsdocs/rfcs/001-agent-mode-architecture.md- Visual architecture diagrams, UI mockups, sequence diagramsImplementation Phases
Discussion Points
This RFC emerged from a discussion about AI agent security in the Clawdbot Discord. Thanks to @devopen for the initial idea and feedback.
cc @sahul_125