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

The description adds substantial behavioral context beyond the sparse annotations: it discloses the one-off install behavior behind invoke, the connect/checkout fallback flows, workspace owner/admin permission requirements for writes, and the installed_in_toolkit vs installed_in_workspace flags on search/describe. This is exactly the kind of non-obvious behavior an agent needs to know. No contradiction with annotations.

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 a single dense paragraph, but every sentence carries distinct information: core flow, invoke vs install, auth/billing edge cases, permission requirements, and the prompt library. It is front-loaded with the key identity ('official mcp.ai marketplace') and uses 'KEY:' to emphasize the most critical behavior. It could be improved with bullet points or separate paragraphs, but for 14 actions and 23 params, it is remarkably compact without fluff.

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 the high complexity (14 actions, 23 optional params, no output schema, minimal annotations), the description covers an impressive amount: the full search→describe→invoke flow, install vs invoke semantics, permission requirements, prompt library sub-usage, and even account-state outcomes (connect/checkout links). However, it omits the 'resume' action entirely and does not specify what search/describe/list_tools return in terms of shape, so an agent is left guessing on those details. Still, it is largely complete for safe 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 0%, so the description must carry the full parameter meaning load. It does explain major params implicitly: action (via the action list), tool_id ('so you pick the right tool_id → invoke RUNS that tool'), prompt_vars (get_prompt fills {{variables}}), prompt_slug/publish_prompt (shareable link), and cancel_reason/cancel_comment (for cancel). However, several params are completely unexplained: resume, immediate, tier_slug, conversation, prompt_tool, message, and report_context. The description gives a good conceptual map but does not fully compensate for the total absence of schema-level descriptions.

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 opens with a specific statement: 'The official mcp.ai marketplace — the in-platform catalog of every MCP/tool, AND the way to run them.' It clearly defines the tool's role as both catalog and executor, and explains its coverage of capability requests ('find an MCP that does X', 'consulta um CPF'). This distinguishes it from siblings like authenticate/connect or specific tools like pref_rj_rio_janeiro_iptu_consultar by positioning it as the meta-tool that discovers and runs such MCPs.

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 when-to-use guidance and exclusions: 'invoke works even when the MCP is NOT installed — it runs the tool pontualmente (one-off)... Use install only to make an MCP PERMANENT... prefer invoke for a single/occasional use.' It also explains the retry flow for auth/payment failures ('invoke returns a connect link... if it is paid and the wallet is empty, invoke returns a checkout/top-up link'), and differentiates sub-actions like 'report_bug sends feedback; request_mcp asks us to build a NEW MCP when nothing fits.'

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