Per-target Layer-1 model card: training counts and held-out scaffold-split metrics (ROC-AUC, PR-AUC, Brier, calibration). Accepts a gene symbol (e.g. EGFR) or UniProt accession (e.g. P00533).
| Name | Required | Description | Default |
|---|---|---|---|
| target | Yes | gene symbol or UniProt accession (required) |
| Name | Required | Description | Default |
|---|---|---|---|
| query_metadata | Yes | Echo of the resolved query. |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
Annotations already declare readOnlyHint, idempotentHint, and destructiveHint false, covering safety. The description adds meaningful context about the output contents (training counts, metrics, calibration) and input formats, which goes beyond the structured annotations without contradicting them.
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 two sentences, front-loaded with the core purpose ('Per-target Layer-1 model card'), and includes essential details without redundancies. Every clause earns its place.
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 a single parameter and an output schema (present), the description adequately covers the tool's behavior and return value semantics. It lacks explicit edge-case handling (e.g., invalid target), but the output schema and simple getter nature make this sufficient.
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 coverage is 100%, so the parameter is fully documented. The description adds concrete examples (EGFR, P00533) and clarifies the two accepted identifier formats, which assists correct invocation beyond the schema's generic 'gene symbol or UniProt accession'.
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?
The description clearly identifies the tool as retrieving a per-target Layer-1 model card with specific metrics (ROC-AUC, PR-AUC, Brier, calibration). It names the resource ('model card') and distinguishes it from sibling tools like get_coverage or get_target_landscape, which serve different purposes.
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 provides clear context on when to use the tool (to fetch model card for a target) and specifies accepted input formats (gene symbol or UniProt accession) with examples. It does not explicitly compare with alternatives or state exclusions, but the intended use is unambiguous.
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.
Most tools have clearly distinct purposes by targeting specific failure types (e.g., ADMET, ADC, bispecific). However, the high number of similarly named 'search_failed_*' and 'search_*_failures' could cause confusion without careful reading, and subtle overlaps exist (e.g., search_failed_adcs vs search_adc_linker_failures are related but distinct).
All tool names follow a consistent verb_noun pattern with snake_case. The verbs 'search_', 'get_', and 'list_' are used appropriately and predictably, with no mixing of conventions.
With 35 tools, the server is on the high side but still appropriate for the broad domain of pharmaceutical failure data across many modalities. Each tool covers a specific niche, though some consolidation might be possible.
The tool surface is remarkably comprehensive, covering failures across small molecules, antibodies, ADCs, bispecifics, PROTACs, oligonucleotides, peptides, vaccines, CRISPR, and more. It includes meta-queries for indicators and targets, leaving no obvious dead ends for agents exploring failure data.
