AI agents use update_draft to create or update resources in Fichapao — usually the action step of a workflow, after the agent has gathered context. Every call changes real data in your Fichapao environment.
This tool modifies existing recipe draft or review versions in the Postgres database. The operation is reversible (can be undone by subsequent updates) and scoped to non-approved versions, preventing direct damage to finalized recipes. It does not execute arbitrary code, delete data, or move financial assets.
From the tool's definition Tool description states 'Atualiza uma versao em rascunho ou em revisao' (Updates a version in draft or in review). The verb 'update' and the stated scope (draft/review versions only, not approved ones) indicate reversible modification of data.
Attacks that exploit this kind of access
Atualiza uma versao em rascunho ou em revisao. Nao altera fichas aprovadas. It is categorised as a Write tool in the Fichapao MCP Server, which means it can create or modify data. Consider rate limits to prevent runaway writes.
Register the Fichapao MCP server in PolicyLayer and add a rule for update_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 Fichapao. Nothing to install.
update_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 update_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 update_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.
update_draft is provided by the Fichapao MCP server (kratos4ai/fichapao-mcp-server). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.
Every MCP server has a record like this.
Type a name, get the same breakdown: verified identity, auth posture, risk grade, capabilities, recommended policy.
Teams ship this data inside their own products. See what a licence covers →