Manage app KV store: config rules (expose/unexpose namespaces) and data-plane operations (get/set/del/incr etc). Actions — Config: - "list_rules": List all KV namespace exposure rules for the app - "expose": Expose a key pattern with read/write role access control - "unexpose": Remove an exposure...
Risk signalsHigh parameter count (21 properties)
Part of the Mcp server.
Free to start. No card required.
AI agents may call manage_kv to permanently remove or destroy resources in Mcp. Without a policy, an autonomous agent could delete critical data in a loop with no way to undo the damage. PolicyLayer blocks destructive tools by default and requires explicit human approval before enabling them.
Without a policy, an AI agent could call manage_kv in a loop, permanently destroying resources in Mcp. There is no undo for destructive operations. PolicyLayer blocks this tool by default and only allows it when a human explicitly approves the action.
Destructive tools permanently remove data. Block by default. Only enable with explicit approval workflows.
{
"version": "1",
"default": "deny",
"hide": [
"manage_kv"
]
}See the full Mcp policy for all 47 tools.
These attack patterns abuse exactly the kind of access manage_kv gives an agent. Each links to the full case and the policy that stops it:
Other destructive tools across the catalogue. The same approach applies to each: deny by default, or require human approval.
Manage app KV store: config rules (expose/unexpose namespaces) and data-plane operations (get/set/del/incr etc). Actions — Config: - "list_rules": List all KV namespace exposure rules for the app - "expose": Expose a key pattern with read/write role access control - "unexpose": Remove an exposure rule by pattern - "stats": Get KV usage stats (key count, memory, etc.) - "scan": Scan keys by prefix (cursor-based, params: prefix, limit, cursor) - "flush": Delete ALL keys in the KV store (requires confirm: true; include_config?: true also wipes rules) Actions — Data plane: - "get": Get the value of a key (opts: raw?, touch?) - "set": Set a key to a value (opts: ttl?, ephemeral?, raw?) - "del": Delete one key - "incr": Increment a key's integer value (opts: by?) - "decr": Decrement a key's integer value (opts: by?) - "setnx": Set a key only if it does not already exist (opts: value, ttl?) - "setex": Set a key with an explicit TTL in seconds (same as set + ttl) - "cas": Compare-and-swap: atomically set next only if current value matches expected - "exists": Check if a key exists - "ttl": Get remaining TTL of a key in seconds - "expire": Set a TTL on an existing key - "mget": Get values of multiple keys at once (uses batch op) - "mset": Set multiple key-value pairs at once (uses batch op; entries: {key: value}) Parameters by action: list_rules: { app_id, action: "list_rules" } expose: { app_id, action: "expose", pattern, read, write } unexpose: { app_id, action: "unexpose", pattern } stats: { app_id, action: "stats" } scan: { app_id, action: "scan", prefix?, limit?, cursor? } flush: { app_id, action: "flush", confirm: true, include_config? } get: { app_id, action: "get", key, raw?, touch? } set: { app_id, action: "set", key, value, ttl?, ephemeral?, raw? } del: { app_id, action: "del", key } incr: { app_id, action: "incr", key, by? } decr: { app_id, action: "decr", key, by? } setnx: { app_id, action: "setnx", key, value, ttl? } setex: { app_id, action: "setex", key, value, ttl } cas: { app_id, action: "cas", key, expected, next } exists: { app_id, action: "exists", key } ttl: { app_id, action: "ttl", key } expire: { app_id, action: "expire", key, ttl } mget: { app_id, action: "mget", keys } mset: { app_id, action: "mset", entries } Warning: "flush" deletes ALL keys and cannot be undone. Always pass confirm: true explicitly.. It is categorised as a Destructive tool in the Mcp MCP Server, which means it can permanently delete or destroy data. Block by default and require explicit approval.
Register the MCP server in PolicyLayer and add a rule for manage_kv: 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 Mcp. Nothing to install.
manage_kv is a Destructive tool with critical risk. Critical-risk tools should be blocked by default and only enabled with explicit human approval.
Yes. Add a rate_limit block to the manage_kv 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 manage_kv. 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.
manage_kv is provided by the MCP server (@butterbase/mcp). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.
Deterministic rules across all 47 Mcp tools. Per-identity grants. Full audit log. Live in minutes. Nothing to install.
Free to start. No card required.
4,600+ MCP servers and 31,000+ tools scanned and risk-classified.