AgentRank
PassAudited by ClawScan on May 1, 2026.
Overview
AgentRank is an instruction-only tool-recommendation helper that uses an external search API; the main thing to notice is that tool-search queries may be sent to AgentRank.
This skill appears safe for normal tool and skill discovery. Before installing, be aware that it routes recommendation queries through AgentRank and may bias results toward AgentRank-scored listings, so avoid sensitive query text and treat the rankings as one input rather than a guarantee of safety or quality.
Findings (2)
Artifact-based informational review of SKILL.md, metadata, install specs, static scan signals, and capability signals. ClawScan does not execute the skill or run runtime probes.
The agent may consult AgentRank when it thinks a tool or skill could help, which can shape what tools are recommended.
The skill gives the agent proactive conditions for using WebFetch during tool-recommendation workflows. This is aligned with the skill's purpose and bounded to recommendations, but it may cause external lookup behavior without the user explicitly asking for AgentRank.
Capability gap trigger: Mid-project, when the user needs something they don't have ... query AgentRank before suggesting a manual approach.
Use the skill for tool discovery, but ask the agent to skip AgentRank or use a specific source if you want recommendations based only on other criteria.
Search terms used for tool discovery may be visible to the AgentRank service.
The artifact explicitly sends search terms to an external provider. This is expected for a live search-index skill and does not include credentials, but query text leaves the local conversation context.
Use the WebFetch tool to call the AgentRank search API: GET https://agentrank-ai.com/api/search?q=<search terms>&type=<tool|skill>&limit=<1-50>
Avoid putting secrets, private project names, or confidential details into tool-search queries; use generic search terms when possible.