Skip to content

Add Clawvisor to Tools & Frameworks#2

Open
clawby-bot wants to merge 1 commit intoLLMSecurity:mainfrom
clawby-bot:add-clawvisor
Open

Add Clawvisor to Tools & Frameworks#2
clawby-bot wants to merge 1 commit intoLLMSecurity:mainfrom
clawby-bot:add-clawvisor

Conversation

@clawby-bot
Copy link
Copy Markdown

What

Adds Clawvisor to the Tools & Frameworks table.

Why it fits

Clawvisor is an open-source AI agent gateway that directly addresses the Privilege Escalation & Excessive Agency and Permission & Access Control attack/defense categories already in this list.

Key security properties:

  • Credential vaulting — agents never hold API keys or OAuth tokens; Clawvisor injects credentials server-side
  • Purpose-based authorization — agents declare task scope upfront, users approve once; Clawvisor enforces on every request
  • Per-request approval queue — any out-of-scope action requires explicit human approval before execution
  • Audit trail — every request, purpose declaration, credential injection, and decision is logged
  • Restrictions — hard blocks on specific service/action pairs (e.g. no agent can delete calendar events)

Works with Claude Code, Claude Desktop (MCP), and any HTTP-based agent.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant