ai_compute_trendYearly training-compute trend from Epoch AI's models database: for each year, the number of models with a compute estimate, the largest training run (FLOP) that year, and which model set it. Optionally filtered by domain and a starting year.
| Name | Required | Description | Default |
|---|---|---|---|
| domain | No | Case-insensitive substring filter on domain, e.g. "Language" | |
| from_year | No | Earliest calendar year to include, default 2010 |
Changes observed during successful MCP inspections.
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
Annotations already declare read-only, idempotent, open-world, non-destructive semantics, so the safety profile is covered. Beyond that, the description adds useful context by naming the data source (Epoch AI's models database) and revealing that only 'models with a compute estimate' are counted, which hints at partial coverage. It omits auth/rate-limit notes, so it is not exhaustive.
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?
Two tightly written sentences: the first front-loads the tool's output contract, the second covers filtering. No filler and nothing buried.
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?
There is no output schema, so the description carries the return-value burden and does it well by enumerating the per-year fields. Parameters are fully covered by the schema and safety semantics by the annotations, leaving no material gap for correct invocation.
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 100%, with domain (case-insensitive substring) and from_year (default 2010) fully documented in the schema. The description merely restates that filtering by domain and start year is optional, adding no syntax or semantics beyond the schema, so the baseline 3 applies.
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?
States a specific aggregate verb (compute trend) over a specific resource (Epoch AI's models database) and enumerates exactly what is produced per year: model count, largest training run in FLOP, and the model that set it. An agent can distinguish this aggregation tool from siblings like ai_models_search or ai_model, which enumerate individual models rather than per-year trends.
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 sentence 'Optionally filtered by domain and a starting year' implies when the tool is useful (getting a yearly compute trend, optionally scoped), but it names no alternatives and states no exclusions or prerequisites. Usage is inferable but not guided.
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.