AI agents use submit_draft to create or update resources in Blind-Auditor-MCP — usually the action step of a workflow, after the agent has gathered context. Every call changes real data in your Blind-Auditor-MCP environment.
This tool creates or modifies data by accepting and recording a code draft submission. It is not destructive (no irreversible deletion), not financial, and does not directly execute code—it stores it for later audit. While the server's purpose involves code auditing and self-correction, submit_draft itself is a Write operation that persists user input rather than executing it.
From the tool's definition The tool 'submit_draft' submits a code draft for audit, which creates or records data in the audit system. The verb 'submit' combined with 'draft' indicates the tool modifies state by storing user-provided code content.
Documented attack patterns abuse exactly the kind of access submit_draft gives an agent:
PolicyLayer is an MCP gateway — it sits between your AI agents and Blind-Auditor-MCP, and nothing reaches the server without passing your rules. This is the rule we recommend for submit_draft:
{
"version": "1",
"default": "deny",
"tools": {
"submit_draft": {
"limits": [
{
"counter": "submit_draft_rate",
"window": "minute",
"max": 30,
"scope": "grant"
}
]
}
}
}submit_draft stays usable, but capped — an agent stuck in a loop can't make hundreds of changes a minute. Everything else on the server is denied unless you say otherwise.
Free to start. No card required.
Submit a code draft for audit. It is categorised as a Write tool in the Blind-Auditor-MCP MCP Server, which means it can create or modify data. Consider rate limits to prevent runaway writes.
Register the Blind-Auditor- MCP server in PolicyLayer and add a rule for submit_draft: 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 Blind-Auditor-MCP. Nothing to install.
submit_draft is a Write tool with medium risk. Write tools should be rate-limited to prevent accidental bulk modifications.
Yes. Add a rate_limit block to the submit_draft 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 submit_draft. 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.
submit_draft is provided by the Blind-Auditor- MCP server (sim-xia/blind-auditor). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.
Start from Blind-Auditor-MCP, add the rest of your stack, and see everything your agents can call. Then put policy on all of it.
Free to start. No card required.
4 Blind-Auditor-MCP tools catalogued and risk-classified — across an index of 43,000+ MCP servers.