Search arXiv for recent papers matching a query (title, authors, abstract, PDF link). Use for ML/AI research agents and literature review.
Example call: {"query": "diffusion transformer"}
Cost: $0.005–$0.05 USDC on Base per call.
| Name | Required | Description | Default |
|---|---|---|---|
| query | Yes |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
No annotations are provided, so the description carries the full burden. It does not disclose behavior such as what happens on no results, pagination, rate limits, or response format beyond a brief list. The cost mention is useful but insufficient for full transparency.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is concise (four lines), front-loaded with purpose, followed by an example and cost. Every sentence is informative, and there is no redundant information.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
Given no output schema, the description lists the fields returned (title, authors, abstract, PDF link) but lacks details like result count, ordering, or recency. It is adequate but could mention that results are limited to recent papers and sorted by date.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
Schema description coverage is 0%, and the description adds only an example call. It does not explain query syntax, operators, or formatting tips. The single parameter 'query' is minimally described beyond its name and example.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
Description clearly states 'Search arXiv for recent papers matching a query' and lists the fields returned (title, authors, abstract, PDF link). This distinguishes it from sibling tools like lookup_wikipedia or lookup_weather, which serve different domains.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
The description explicitly says 'Use for ML/AI research agents and literature review', providing a clear context of use. However, it does not mention when not to use it or compare to alternatives among the many sibling lookup tools.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Add one secure layer between your agents and this server.
The set is riddled with near-duplicates: lookup_reddit/scrape_reddit, lookup_wikipedia/scrape_wikipedia, lookup_dockerhub/scrape_dockerhub, lookup_steam/scrape_steam, enrich_googlereviews/enrich_reviews, lookup_ip/lookup_ipinfo, and multiple crypto-pricing tools (lookup_crypto, lookup_coingecko, bundle_crypto_360, scrape_binance, scrape_coinbase). Descriptions try to differentiate with phrases like 'heavier than' or 'same domain but with full thread parsing,' but the boundaries are fuzzy and an agent can easily pick the wrong one.
Naming follows a fairly consistent prefix-based snake_case pattern (lookup_, scrape_, enrich_, bundle_, search_, ai_, data_) where the prefix denotes action weight and the noun identifies the target. Minor deviations exist: posts_x, ai_ask/pro/ultra (model-tier names instead of resources), sslstatus (missing underscore), and lookup_useragents_top are slightly off-pattern.
172 tools is an extreme count, far beyond even the 50+ floor for a score of 1. This floods the agent's context and tool-selection space, making every call require a search through a massive list. While aggregation servers can justify more tools, this volume is unmanageable and every tool must be evaluated by the agent.
The surface is extraordinarily broad but unevenly deep: many sources have both a light lookup and a heavy scrape variant, while other areas have just a single shallow endpoint. There is no coherent domain with complete lifecycle coverage, and despite the huge catalog, common capabilities are still absent. The breadth prevents obvious gaps, but depth and coherence suffer.