record_predictionsRecord a complete structured prediction document for downstream evaluation. Call once with the exact machine-readable prediction; provide the human-readable report separately.
| Name | Required | Description | Default |
|---|---|---|---|
| predictions | Yes | The complete structured prediction document for this run. |
| Name | Required | Description | Default |
|---|---|---|---|
| recorded | Yes |
Changes observed during successful MCP inspections. Dates show when Glama detected each change.
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
The description says 'record,' which strongly implies a write operation, while the annotation readOnlyHint=true declares the operation as read-only. This is a direct contradiction. No other behavioral context is provided.
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 sentences, front-loaded with the purpose, then usage instruction. Every word is necessary, with no filler or repetition.
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?
The description gives basic usage but leaves the side-effect ambiguity unresolved due to the contradiction between 'record' and readOnlyHint. A single nested object parameter and an output schema exist, yet the description fails to clarify whether or how data is persisted, making it incomplete.
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 already describes the 'predictions' parameter with 100% coverage. The description adds meaningful format guidance: 'exact machine-readable prediction' and separating the human-readable report, which enriches the parameter's intended use beyond the schema.
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 uses a specific verb ('record') with a clear object ('complete structured prediction document') and purpose ('for downstream evaluation'). It clearly distinguishes this tool from siblings, which are mostly read-only getters or calculations.
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 gives explicit usage instructions: call once with the exact machine-readable prediction, and provide the human-readable report separately. It does not name alternative tools, but no sibling appears to serve the same recording function, making context clear.
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 target distinct data sources, but several near-duplicates exist: get_token_approvals/get_wallet_approvals, get_defi_positions/get_wallet_positions, and get_wallet_portfolio/get_eth_balance. The descriptions do cross-reference and clarify the differences, so an agent can disambiguate with effort, but names alone are not enough.
The dominant get_<noun> pattern is clear and nearly all names use lowercase snake_case with verb-first conventions. A few tools like calculate, record_predictions, http_fetch, and web_search break the get_ pattern, but the overall style remains predictable.
34 tools is excessive for a single server, even for a broad DeFi/onchain analytics domain. The count is inflated by generic utilities such as calculate, count_text_stats, web_search, and http_fetch, plus multiple overlapping data-retrieval endpoints, making the surface hard to scan.
The set covers an unusually wide range of domain operations: prices, balances, portfolio/positions, approvals, yields, TVL, DEX quotes/volume, transactions, blocks, gas, ENS, contract reads, and risk assessments. It is view-only by design, so missing write/transaction tools is acceptable; minor gaps like address-based token pricing or transaction simulation are workaround-able.