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

Annotations already cover readOnly=true, destructive=false and idempotent=true, and the description reinforces this usefully with 'opens nothing by itself', so an agent knows no mutation occurs in Vtiger. It also adds a concrete picture of the return payload (screens, buttons, fields walked through step by step), which annotations do not convey. It stops short of describing the shape or length of the returned instructions or any failure mode when the Cloud Browser is unavailable.

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 critical fact—that this returns instructions and performs nothing—is buried in the second sentence behind an opening line that implies the opposite. 'From ChatGPT, Claude or any AI agent' is promotional filler that earns no place, and the API rate-limit sentence is a dense run-on that could be compressed to a clause. The rationale is genuinely useful, but the front-loading is wrong for a tool whose main risk is being mistaken for a data fetch.

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?

With no parameters and no output schema, the description carries the burden of explaining what comes back, and it does: a step-by-step walkthrough of screens, buttons and fields, plus the follow-up path through the browser tools. Remaining gaps are modest for a zero-argument tool—no guidance on prerequisites such as the Vtiger Cloud Browser session state and no statement of how complete or account-specific the returned guide is.

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?

The tool takes zero parameters, so per the baseline this dimension cannot be penalized; there is no argument semantics for the description to explain. The absence of any parameter also matches the stated behavior of returning a static procedural guide rather than querying a filtered record set.

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 eventually makes the actual function explicit: it 'Returns how to do it in Vtiger step by step (screens, buttons, fields) and opens nothing by itself', which is a specific and unusual purpose an agent could otherwise get wrong. The opening sentence 'List deals in the user's own Vtiger account' reads like the tool performs the listing, so the name/title framing and the real behavior pull in opposite directions, and the correction only arrives in sentence two. It does differentiate from the cotlet_browser_* siblings by assigning them the actual execution, but says nothing about how it relates to vtiger_list_contacts or vtiger_list_cases.

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?

It gives clear routing: take the returned step-by-step procedure and then execute it 'with the cotlet_browser_* tools in the user's Vtiger Cloud Browser', which tells the agent what to do next rather than leaving it to infer. The paragraph on Vtiger's REST API and rate limits (0 for One Pilot, 24,000/day at 30/min for One Growth, no official MCP server) explains why this guide-plus-browser path exists instead of a direct API read. What is missing is an explicit exclusion, e.g. when an agent should not call this at all (already has data, user not signed into Cloud Browser).

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