Report the active demo wallet (live mode), confirm default demo mode (no wallet set), or report why demo mode isn't active when the env var is missing or misconfigured. ALWAYS enumerates the available personas + their addresses + descriptions regardless of VAULTPILOT_DEMO state, so the agent can ...
AI agents call get_demo_wallet to retrieve information from VaultPilot MCP without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
Even though get_demo_wallet only reads data, uncontrolled read access leaks sensitive information and racks up API costs — an agent caught in a retry loop can make thousands of calls a minute without anyone noticing.
Attacks that exploit this kind of access
Report the active demo wallet (live mode), confirm default demo mode (no wallet set), or report why demo mode isn't active when the env var is missing or misconfigured. ALWAYS enumerates the available personas + their addresses + descriptions regardless of VAULTPILOT_DEMO state, so the agent can offer the user a choice without hardcoding the list (issue #392). RESPONSE: { demoActive, mode, envState: 'enabled' | 'unset' | 'invalid', personas, matrix, [active], [message] }. When envState is 'unset' or 'invalid' the message field tells the user how to fix it (set VAULTPILOT_DEMO=true exact literal, lowercase). When envState is 'enabled', active carries the current live wallet (or null in default demo mode). MATRIX (issue #409 enrichment): each cell in matrix[chain][type] carries a rehearsableFlows: string[] field listing multi-step / state-dependent demo flows the cell's existing on-chain state already supports end-to-end (e.g. aave_supply, marinade_stake, swap_eth_usdc), and an optional flowGaps field listing flows the persona archetype implies but the wallet's current state does NOT support, with a one-line recommendation (switch persona / exit demo / pair Ledger). Read these BEFORE the user picks a flow so you offer rehearsable flows up-front and steer them off flows that would loop on missing on-chain state (the simulated send returned in demo mode never mutates state, so any flow whose precondition is itself a state change can't be rehearsed end-to-end against a wallet that doesn't already have it). It is categorised as a Read tool in the VaultPilot MCP MCP Server, which means it retrieves data without modifying state.
Register the VaultPilot MCP server in PolicyLayer and add a rule for get_demo_wallet: 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 VaultPilot MCP. Nothing to install.
get_demo_wallet 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 get_demo_wallet 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 get_demo_wallet. 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.
get_demo_wallet is provided by the VaultPilot MCP server (vaultpilot-mcp). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.