Zanna Aperta

MCP Bridge completo per OpenClaw con 45 tool per agenti, workspace, progetti, cron, browser, canvas, nodes, messaging, gateway, Ollama e ClawX

Security Scan

The skill's code, instructions, and requested access are consistent with a local MCP/OpenClaw bridge: it manipulates OpenClaw workspaces and invokes the OpenClaw CLI, and it does not request unrelated credentials or external downloads.

Purpose & Capability

Name/description claim (MCP Bridge for OpenClaw exposing many tools) matches the code and SKILL.md: the script runs an OpenClaw CLI (OPENCLAW_BIN), manipulates workspace directories, and implements the listed tool handlers (agents, workspace, projects, cron, browser, exec, Ollama, ClawX, etc.). Nothing required by the code (workspace access, openclaw binary, optional local Ollama/ClawX) is out of scope for a bridge tool.

Instruction Scope

SKILL.md instructs adding the Python script to ~/.openfang/settings.json and to set OPENCLAW_WORKSPACE/OPENCLAW_BIN — the script reads/writes workspace files and uses the OpenClaw CLI. The instructions do not ask for unrelated files, secrets, or external endpoints beyond optional local services (ollama on localhost and ClawX path). The code does perform filesystem I/O within the workspace and spawns OpenClaw/other local commands, which is expected behavior for this functionality.

Install Mechanism

There is no external install spec or network download; the skill is instruction-only and ships a Python file. Installation consists of placing and running the included script in the user's OpenClaw skills path — no remote archives or third‑party package installs are performed by the skill itself.

Credentials

The skill requests no credentials and the only environment/config it uses (OPENCLAW_WORKSPACE, OPENCLAW_BIN) is directly relevant to its purpose. It will read and write files in the workspace directories it manages; this file access is proportionate to a workspace/agent management bridge but users should note any secrets stored in those directories could be accessed.

Persistence & Privilege

always is false and model invocation is allowed (platform default). The skill does not request permanent platform-level privileges or modify other skills' configurations. It will create, update, and delete files under the configured workspace and may start/stop local services (ClawX) which is consistent with its role.

Assessment

This skill appears coherent for a local MCP/OpenClaw bridge, but it executes local commands and reads/writes your OpenClaw workspace. Before installing: (1) review the included zanna-aperta.py yourself (it will run with your user permissions); (2) ensure no sensitive credentials or secrets are stored under the configured workspace path; (3) be aware it can create/delete workspaces, invoke OpenClaw tools, and interact with local services (Ollama on localhost and ClawX); (4) run it in a test or isolated environment if you want to limit risk; and (5) verify the openclaw binary and any local services it talks to are trusted. If you want extra assurance, ask for a full line-by-line review of the Python file (the provided snippet is large) or run the skill with a read-only workspace copy first.

MCP Bridge che collega OpenFang (o qualsiasi client MCP) a tutti i tool OpenClaw nativi + ClawX.

Tool Disponibili (45)

Agenti (3): agent_find, agent_update, agent_delete

Workspace (5): workspace_create, workspace_list, workspace_find, workspace_update, workspace_delete

Progetti (5): project_create, project_list, project_find, project_update, project_delete

Cron (4): cron_list, cron_add, cron_remove, cron_run

Browser (4): browser_open, browser_snapshot, browser_click, browser_type

Canvas (3): canvas_present, canvas_hide, canvas_eval

Nodes (3): nodes_list, nodes_notify, nodes_camera_snap

Messaging (2): message_send, message_poll

Gateway (2): gateway_status, gateway_restart

Exec (2): exec_docker, exec_git

Ollama (4): ollama_list, ollama_chat, ollama_generate, ollama_pull

ClawX (4): clawx_start, clawx_stop, clawx_status, clawx_restart

Configurazione

Aggiungi a ~/.openfang/settings.json:

{
  "mcpServers": {
    "zanna-aperta": {
      "command": "python3",
      "args": ["~/.openclaw/skills/zanna-aperta/zanna-aperta.py"],
      "env": {
        "OPENCLAW_WORKSPACE": "~/.openclaw/workspace",
        "OPENCLAW_BIN": "openclaw"
      }
    }
  }
}

Requisiti

• OpenClaw installato
• Python 3.8+
• Ollama in esecuzione su localhost:11434 (opzionale)
• ClawX in ~/.openclaw/workspace/ClawX (opzionale)

Autore

Raid AI - 2026