# create_agent

Create a new autonomous strategy agent. Gets a funded wallet (500 bsUSD) and API key. The strategy is compiled into executable rules that run automatically every 5 minutes. Credentials are saved locally to ~/.conviction/agents.json. Limit: 10 active agents per owner, 3 new agents per hour.

Agent View of the PolicyLayer registry record for `create_agent`. HTML page: https://policylayer.com/tools/abcxz-conviction-fm/create-agent

## Facts

- Tool: `create_agent`
- Server: Conviction (`abcxz/conviction-fm`) — https://policylayer.com/tools/abcxz-conviction-fm.md
- Homepage: https://github.com/abcxz/conviction-fm
- Risk category: Financial (Critical risk)
- Registry record: grade D, identity unverified
- Server rate-limited: no
- Parameters: 0
- Recommended policy verdict: Approval-gated

## Example call (MCP tools/call, JSON-RPC 2.0)

```json
{
  "jsonrpc": "2.0",
  "id": 1,
  "method": "tools/call",
  "params": {
    "name": "create_agent",
    "arguments": {}
  }
}
```

## Why create_agent is rated Critical

This tool creates autonomous financial agents that receive real funded wallets (500 bsUSD), execute strategies automatically, and participate in competition pools involving financial positions. The financial implications are direct and significant — misuse could result in deploying multiple money-managing agents that autonomously open/close positions and commit funds without further human oversight.

From the tool's own definition: "Gets a funded wallet (500 bsUSD) and API key... strategy agent... run automatically every 5 minutes... Limit: 10 active agents per owner"

Risk signals: Bulk/mass operation — affects multiple targets

## Use case

AI agents use create_agent to commit financial operations through Conviction, usually the final step of a payment, billing, or trading workflow. A call moves real money.

## Recommended policy (PolicyLayer)

Verdict: **Approval-gated**. Enforced by the PolicyLayer MCP gateway (https://policylayer.com/mcp-gateway) before a call reaches Conviction:

```json
{
  "version": "1",
  "default": "deny",
  "tools": {
    "create_agent": {
      "deny_if": [
        {
          "conditions": [],
          "on_deny": "Requires human approval."
        }
      ]
    }
  }
}
```

## Other tools on Conviction (9)

- `toggle_agent` — Execute — https://policylayer.com/tools/abcxz-conviction-fm/toggle-agent.md
- `enter_position` — Financial — https://policylayer.com/tools/abcxz-conviction-fm/enter-position.md
- `export_wallet` — Financial — https://policylayer.com/tools/abcxz-conviction-fm/export-wallet.md
- `get_agent_status` — Read — https://policylayer.com/tools/abcxz-conviction-fm/get-agent-status.md
- `get_leaderboard` — Read — https://policylayer.com/tools/abcxz-conviction-fm/get-leaderboard.md
- `get_my_agents` — Read — https://policylayer.com/tools/abcxz-conviction-fm/get-my-agents.md
- `get_pool_history` — Read — https://policylayer.com/tools/abcxz-conviction-fm/get-pool-history.md
- `get_pools` — Read — https://policylayer.com/tools/abcxz-conviction-fm/get-pools.md
- `update_strategy` — Write — https://policylayer.com/tools/abcxz-conviction-fm/update-strategy.md

## For agents

This record is a snapshot. Live verdicts and the full registry:

- Check every server in your MCP config at once: `npx -y policylayer stack`
- Vet a server before you add it: install the mcp-precheck skill — `npx skills add https://policylayer.com` (skill text: https://policylayer.com/skill.md)
- Query the registry over MCP: endpoint `https://api.policylayer.com/mcp` — tools `check_mcp_server`, `check_mcp_stack`, `check_tool`, `search_registry`, `get_change_events`

---

Source: the PolicyLayer MCP registry — one continuously verified record per MCP server. Full record: https://policylayer.com/registry?q=abcxz-conviction-fm · API: https://policylayer.com/registry/api · Policy library: https://policylayer.com/policies/abcxz-conviction-fm