AI agents call inspect_local_extension to retrieve information from Todos without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
The tool exclusively validates and inspects existing local files and manifests. The explicit phrase 'without installing it' confirms no state-changing actions occur. This is a diagnostic/query operation typical of Read category tools (inspect, validate, check). The blast radius is minimal since it cannot modify data, execute code, or alter system state.
From the tool's definition Tool performs validation and inspection of local extension manifests/directories/bundles 'without installing it' — a pure read-only operation with no side effects or modifications.
Attacks that exploit this kind of access
Validate a local extension manifest, directory, or offline bundle without installing it. It is categorised as a Read tool in the Todos MCP Server, which means it retrieves data without modifying state.
Register the Todos MCP server in PolicyLayer and add a rule for inspect_local_extension: 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 Todos. Nothing to install.
inspect_local_extension is a Read tool with low risk. Read-only tools are generally safe to allow by default.
Yes. Add a rate_limit block to the inspect_local_extension 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 inspect_local_extension. 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.
inspect_local_extension is provided by the Todos MCP server (@hasna/todos). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.
Every MCP server has a record like this.
Type a name, get the same breakdown: verified identity, auth posture, risk grade, capabilities, recommended policy.
Teams ship this data inside their own products. See what a licence covers →