// SCAN REPORT

AgentOS

35 tools. 12 can modify or destroy data without limits.

1 destructive tool with no built-in limits. Policy required.

Last updated: 3 Jul 2026

12 can modify or destroy data

23 read-only

35 tools total

Community server · catalogue entry verified 03/07/2026

How to control AgentOS ↓

TOOL MAP

What AgentOS exposes to your agents

Read (23) Write / Execute (11) Destructive / Financial (1)

Running more than one server? Check your whole stack →

MOST DANGEROUS TOOLS

Critical Risk

The most dangerous AgentOS tools

Destructive · High kv_delete This tool permanently deletes data from persistent memory without any undo capability. Execute · Medium template_render Template rendering is more than a passive read; it executes template logic with user-supplied variables, which can include arbitrary expressions or code depending on the templating engine. Write · Medium budget_set This tool creates or modifies budget limit settings, which is a reversible write operation. Write · Medium kv_set The tool stores data in persistent key-value memory, which is a create/modify operation.

12 of AgentOS's 35 tools can modify, destroy, or commit something on every call — and an agent calls them with no built-in limits.

POLICY

How to control AgentOS

PolicyLayer is an MCP gateway — it sits between your AI agents and AgentOS, and nothing reaches the server without passing your rules. These are the rules we recommend:

Deny destructive operations

{
  "kv_delete": {
    "deny_if": [
      {
        "conditions": [],
        "on_deny": "Blocked by default. Requires approval."
      }
    ]
  }
}

Destructive tools should never be available to autonomous agents without human approval.

Rate limit write operations

{
  "budget_set": {
    "limits": [
      {
        "counter": "budget_set_per_hour",
        "window": "hour",
        "max": 30,
        "scope": "grant"
      }
    ]
  }
}

Prevents bulk unintended modifications from agents caught in loops.

Cap read operations

{
  "analytics_dashboard": {
    "limits": [
      {
        "counter": "analytics_dashboard_per_minute",
        "window": "minute",
        "max": 60,
        "scope": "grant"
      }
    ]
  }
}

Controls API costs and prevents retry loops from exhausting upstream rate limits.

Create a free account and register AgentOS — nothing to install.
Add these rules — paste them, or build them visually. Tune the limits to your setup.
Point your MCP client (Claude, Cursor, anything) at your gateway URL.

ENFORCE POLICY ON AGENTOS →

Instant setup, no code required.

FULL CATALOGUE

All 35 AgentOS tools

DESTRUCTIVE 1 tools

Destructive kv_delete Remove a key-value pair from persistent local memory.

EXECUTE 1 tools

Execute template_render Render a template with provided variables.

WRITE 10 tools

Write budget_set Set a budget limit for a session or task to track resource usage. Write format_convert Convert between file formats locally. Supports MD↔HTML, CSV↔JSON, YAML↔JSON, XML→JSON. Write kv_set Store a key-value pair in persistent local memory with optional TTL and namespace. Write log_action Append a structured log entry for debugging and audit trail. Write monitoring_record_metric Record a performance or business metric. Write support_create_ticket Create a new support ticket for issues, questions, or feature requests. Write task_checkpoint Save a checkpoint for an active task to preserve current state. Write task_complete Mark a task as completed with final state. Write task_create Create a new task with initial state for tracking progress across sessions. Write template_create Create a new custom template.

READ 23 tools

Read analytics_dashboard Get usage analytics dashboard for today or specified date range. Read analytics_report Generate detailed usage report for a date range. Read budget_check Check current budget usage and remaining limits. Read budget_consume Record resource consumption and check if budget is exceeded. Auto-detects loops. Read kv_get Retrieve a value by key from persistent local memory. Read kv_list List keys in a namespace with optional prefix filter. Read log_export Export logs as JSON or CSV for analysis. Read log_search Search logs by tool name, date range, or keyword. Read log_summary Get a summary of log activity for a time period. Read monitoring_acknowledge_alert Acknowledge an active alert. Read monitoring_business_report Generate business intelligence reports. Read monitoring_capacity_analysis Analyze current capacity and get scaling recommendations. Read monitoring_get_alerts Get active alerts and manage alert status. Read monitoring_get_metrics Get current metrics and historical data. Read monitoring_system_status Get overall system health and monitoring overview. Read support_admin_stats Get support system statistics (admin only). Read support_knowledge_base Search the knowledge base for help articles and FAQs. Read support_list_tickets List all support tickets for a user. Read support_ticket_status Check the status of a support ticket and view responses. Read task_list List all tasks for a session with optional status filter. Read task_rollback Rollback a task to a previous checkpoint. Read task_status Get current status and state of a task. Read template_list List all available templates with optional category filter.

EXPLORE

Related servers

Other MCP servers with similar tools — same risk classification, starter policies for each.

FAQ

Questions about AgentOS

Can an AI agent delete data through the AgentOS MCP server? +

Yes. The AgentOS server exposes 1 destructive tools including kv_delete. These permanently remove resources with no undo. PolicyLayer blocks destructive tools by default so they never reach the upstream server.

How do I prevent bulk modifications through AgentOS? +

The AgentOS server has 10 write tools including budget_set, format_convert, kv_set. Set a rate limit in your policy -- for example, 10 calls per hour prevents an agent from making more than 10 modifications per hour. PolicyLayer enforces this at the gateway, before calls reach AgentOS.

How many tools does the AgentOS MCP server expose? +

35 tools across 3 categories: Destructive, Read, Write. 23 are read-only. 12 can modify, create, or delete data.

How do I enforce a policy on AgentOS? +

Register the AgentOS MCP server in PolicyLayer, apply the suggested rules above (adjust the limits to your use case), and point your AI client at the PolicyLayer proxy URL instead of the server directly. Your agents keep the same tools; PolicyLayer evaluates every call against policy before it executes. Nothing to install, live in minutes.

Enforce policy on every AgentOS tool call.

Deterministic rules across all 35 AgentOS tools. Per-identity grants. Full audit log. Live in minutes. Nothing to install.

ENFORCE POLICY ON AGENTOS →

Instant setup, no code required.

35 AgentOS tools catalogued and risk-classified — across an index of 43,000+ MCP servers.