Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

With no annotations supplied, the description carries the full burden and does so: sandbox capabilities (await, console.log, return), where the return value lands (`resultado`), hard limits (100 calls, 15 s), the cost model (1 credit per execution plus the sum of internal calls), and explicit negative capability ("Sin acceso a red, archivos ni módulos"). That is precisely the risk-relevant information an agent needs before writing arbitrary code.

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?

Purpose is front-loaded in the first clause, followed by use cases, execution semantics, cost/limits, and finally the callable surface. The enumerated function catalog is long but earns its place because an agent literally cannot write valid code without knowing the `api.*` names, signatures and prices; a single minor oddity is the stray trailing "Costo: 1 crédito(s)" fragment at the end.

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?

For a high-complexity execution tool with no annotations and no output schema, the description covers execution model, limits, cost, return channel and the available functions. The remaining gap is failure semantics: it does not say whether an exception inside the code still costs a credit, what happens on timeout mid-batch, or how per-internal-call errors appear in `resultado`.

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 baseline is 3, but the description adds real meaning the schema does not: the `codigo` string runs against a pre-authenticated `api` object, may use top-level await and console.log, and its return value surfaces as `resultado`. The 15 s ceiling also contextualizes `timeout_ms`, though it never explains how per-call timeouts interact with the overall wall clock.

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?

"Corre JavaScript en un sandbox con un objeto `api` ya autenticado" states a specific verb (execute) and a specific mechanism (sandboxed JS over an authenticated API proxy), and immediately clarifies the relationship to the sibling tools: one async function per tool with the same parameters and result as calling them directly. An agent can distinguish code mode from a direct single call without opening any schema.

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?

Explicitly names the use cases that select this tool: batches ("validar 50 CBUs, armar 20 cronogramas") and chaining calls without returning to the model, which is exactly the trade-off versus the single-call siblings. It stops short of stating the inverse rule ("for a single check, use validar_cbu directly"), so the agent must infer when-not, but the intended context is clear.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.