ssh_connect

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

With no annotations provided, the description carries the behavioral burden. It explicitly discloses that for an unknown server, credentials are NOT saved ('доступы НЕ сохраняются') and that a next_action will require asking the user for a name and calling ssh_save_server. It also warns not to fabricate a name. This is valuable behavioral context that prevents the agent from inadvertently persisting credentials or guessing server names. However, it does not describe the success response or side effects (e.g., whether a session is kept open), so it's not fully comprehensive.

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 concise and front-loaded. The first sentence states the primary purpose. The second sentence delivers the critical edge case (unknown server) and an explicit instruction to the agent. There is no filler or repetition; every sentence earns its place. The formatting is clear and easy to parse.

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 covers the main flow and the unknown-server edge case, but it omits details about the successful connection outcome (e.g., what happens after connection, whether a session object is returned) since there is no output schema. It also does not explain parameters like port and probe, which have no schema descriptions. Given the tool has 8 parameters and no annotations, the description is not fully sufficient for an agent to invoke it correctly in all scenarios, but it provides the most critical guidance.

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 only 25% (2 of 8 parameters have descriptions in the schema). The description compensates partially by mentioning 'логином и паролем/ключом' (login and password/key), which maps to user, password, key_file, and passphrase. However, it does not clarify the purpose of port or probe, nor does it elaborate on the save_as parameter beyond what the schema says. Given the low coverage, the description should offer more parameter guidance, but the authentication context is a useful contribution.

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 states the tool's function: 'Подключиться к серверу по имени либо по адресу с логином и паролем/ключом' – connect to a server by name or address with login and password/key. It specifies the resource (server) and the action (connect), and distinguishes it from sibling tools like ssh_run or ssh_upload by its focus on establishing a connection. The unknown-server branch is also described, making the purpose unambiguous.

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 for when to use this tool: when a user wants to connect to a server. It also instructs the agent on handling an unknown server: do not invent a name, and expect a next_action to prompt for a name and call ssh_save_server. While it doesn't explicitly mention alternative tools (e.g., ssh_shell_open for interactive shell), the condition of unknown server gives practical guidance. A full exclusion list is not necessary given the tool's self-explanatory name.

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