Detect slop patterns using both fast regex/NLP detectors and semantic LLM analysis via MCP sampling. Catches all 48 rules including subtle patterns like throat-clearing, grandiose stakes, and fractal summaries that regex cannot detect. Slower than detect_slop due to the sampling call.
AI agents call detect_slop_full to retrieve information from Slop Cop without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
This tool retrieves and analyzes information about writing patterns without modifying data, executing code, destroying information, or making financial transactions. The 'full' variant simply uses both regex/NLP and semantic analysis methods to audit prose quality, making it a read-only analysis tool. The lower performance due to sampling calls does not change its fundamental Read category.
From the tool's definition Tool description states it "Detect[s] slop patterns" and performs "analysis" via LLM sampling. The verb is purely analytical/detective with no modification, creation, deletion, or execution of external operations.
Attacks that exploit this kind of access
Detect slop patterns using both fast regex/NLP detectors and semantic LLM analysis via MCP sampling. Catches all 48 rules including subtle patterns like throat-clearing, grandiose stakes, and fractal summaries that regex cannot detect. Slower than detect_slop due to the sampling call. It is categorised as a Read tool in the Slop Cop MCP Server, which means it retrieves data without modifying state.
Register the Slop Cop MCP server in PolicyLayer and add a rule for detect_slop_full: 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 Slop Cop. Nothing to install.
detect_slop_full is a Read tool with low risk. Read-only tools are generally safe to allow by default.
Yes. Add a rate_limit block to the detect_slop_full 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 detect_slop_full. 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.
detect_slop_full is provided by the Slop Cop MCP server (0ju1c3/slop-cop). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.
detect_slop_full is one line of Slop Cop's registry record.
The record carries the whole server: verified identity, auth posture, risk grade, every tool classified, recommended policy — re-checked continuously.
Teams ship this data inside their own products. See what a licence covers →