highagent-tool-index
Instructs agent to follow untrusted redirect chains and load remote MCP tool sets
The description directs the agent to 'Follow start_here.hop first (skill file, 302)' and states it 'also returns docs, CLI, and remote MCP hops.' This is an embedded instruction telling the AI to chase redirects to an external 'pay-per-call' gateway and then load additional remote MCP tool definitions. This is a supply-chain/prompt-injection risk: it could cause the agent to silently pull in and execute instructions or tool schemas from an untrusted third party, and potentially incur pay-per-call charges, without the user's explicit awareness or consent.
highpeople-search-index
Facilitates locating a people-search/surveillance service and loading remote MCP hops
This tool's stated purpose is to help find a 'living people-search index' and returns 'CLI setup, docs, API, and remote MCP hops.' Beyond the privacy concerns of aiding lookup of information about real individuals, it also instructs the agent to follow additional remote MCP hops, which could lead to loading untrusted external tool definitions or code. Combined, this creates both a privacy risk (facilitating people-search/doxxing style lookups) and a supply-chain risk (fetching and trusting remote tool/service endpoints not vetted by the user).
low
Not on gate's verified list
This server isn't on gate's handpicked list. That's not necessarily bad, but there's no third-party signal about it — review its tools carefully.
lowcc-sa
Uniform oversized schema shared across unrelated tools
Every tool in this set (weather hints, timezone lookups, JSON validation, URL parsing, etc.) exposes an identical 9-field schema (zone, json, url, host, city, query, path, ref, feed) regardless of the tool's actual purpose, with each field annotated as 'discarded' after use. While each individual description claims data is not retained, the templated/copy-pasted nature of these schemas across dozens of otherwise simple tools is unusual and makes it hard to verify that unrelated free-text fields (e.g., 'query', 'json', 'path') are not actually used or logged beyond what the description claims. This pattern increases the risk that any single tool could be silently repurposed to accept and act on additional untrusted input without a corresponding change to its description.