AI agents invoke reconnect to trigger actions in Frida Agent MCP. What it does depends on the arguments the agent supplies, and its effects often reach beyond the immediate call — builds kicked off, notifications sent, workflows started.
Reconnecting to a Frida gadget re-establishes an instrumentation session on a running Android process, triggering external operations against a live process. While it sounds benign, it re-enables dynamic analysis/injection capabilities and interacts with external process state.
From the tool's definition Reconnect to the gadget after an app crash
Documented attack patterns abuse exactly the kind of access reconnect gives an agent:
PolicyLayer is an MCP gateway — it sits between your AI agents and Frida Agent MCP, and nothing reaches the server without passing your rules. This is the rule we recommend for reconnect:
{
"version": "1",
"default": "deny",
"tools": {
"reconnect": {
"limits": [
{
"counter": "reconnect_rate",
"window": "minute",
"max": 10,
"scope": "grant"
}
]
}
}
}reconnect stays usable, but rate-capped — a runaway agent can't fire it dozens of times a minute. Everything else on the server is denied unless you say otherwise.
Free to start. No card required.
Reconnect to the gadget after an app crash. It is categorised as a Execute tool in the Frida Agent MCP MCP Server, which means it can trigger actions or run processes. Use rate limits and argument validation.
Register the Frida Agent MCP server in PolicyLayer and add a rule for reconnect: allow, deny, rate-limit, or require approval. Point your MCP client at the PolicyLayer proxy URL and the rule is enforced on every call, before it reaches Frida Agent MCP. Nothing to install.
reconnect is a Execute tool with high risk. Execute tools should be rate-limited and have argument validation enabled.
Yes. Add a rate_limit block to the reconnect rule in your PolicyLayer policy. For example, setting max: 10 and window: 60 limits the tool to 10 calls per minute. Rate limits are tracked per agent session and reset automatically.
Set action: deny in the PolicyLayer policy for reconnect. The AI agent will receive a policy violation error and cannot call the tool. You can also include a reason field to explain why the tool is blocked.
reconnect is provided by the Frida Agent MCP server (1193776794/frida-mcp). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.
Deterministic rules across all 10 Frida Agent MCP tools. Per-identity grants. Full audit log. Live in minutes. Nothing to install.
Free to start. No card required.
10 Frida Agent MCP tools catalogued and risk-classified — across an index of 42,500+ MCP servers.