registry / @noclaf/mcp / 0.3.3

@noclaf/mcp@0.3.3

Package no longer supported. Contact Support at https://www.npmjs.com/support for more info.

MCP server e CLI pra controlar o NOS (Noclaf Operation System) a partir de assistentes de IA como Claude Code, Codex CLI e GitHub Copilot.

AI Security Review

scanned 19d ago · by lpm-firewall-ai

No confirmed malicious attack surface; network, credential, file-write, and child-process behavior are aligned with an MCP/CLI for Noclaf task and AI-harness sync. The main risk is intentional remote-approved vault content being written to AI client surfaces after user setup.

Static reason

One or more suspicious static signals were detected.; previous stored version diff introduced dangerous source

Trigger

User runs noclaf login/init/sync/pull/install-client or starts configured noclaf-mcp server.

Impact

Can configure MCP clients, store a Noclaf PAT locally, call Noclaf/Supabase APIs, and sync approved commands/skills/prompts to Claude or Codex directories.

Mechanism

authenticated Supabase MCP client plus AI-harness provisioning

Rationale

Static inspection found powerful but disclosed and package-aligned MCP/CLI behavior, with no install-time execution, credential harvesting beyond user-provided Noclaf PAT use, arbitrary shell execution, persistence, or destructive behavior. Remote AI-harness sync is a real control-surface capability but is the advertised product function and is scoped through approved vault docs and manifest-based writes.

Evidence

package.jsonREADME.mddist/auth.jsdist/constants.jsdist/supabase.jsdist/mcp-server.jsdist/lib/provision.jsdist/lib/dependencies.jsdist/lib/client-configs.jsdist/lib/vault-source.jsdist/lib/targets.jsbundle/README.md~/.noclaf/credentials.json

@noclaf/mcp@0.3.3

AI Security Review

Decision evidence

Source & flagged code

Source & flagged code

Findings

Source & flagged code