session_start

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

Annotations already declare readOnly=false, destructive=true, openWorld=true, so the safety profile is known. The description adds genuinely non-schema context: execution is local ('Runs on this machine') and desktop actions are serialized (a concurrency guarantee). It omits what a started session consumes, whether sessions leak without session_close, and what the call returns.

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?

Three short, front-loaded sentences with no filler; the launch/attach split comes first and the execution constraints follow. Brevity is close to the failure line given the schema complexity, but nothing is wasted.

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?

A tool with a single required nested object holding thirteen undocumented options, destructive semantics, and no output schema needs far more than three sentences. An agent gets no idea how to choose a kind/channel/browser combination or what happens to the process afterwards.

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% across 13 nested target properties, so the description must carry the load. It only gestures at cdpUrl and pid ('CDP endpoint/native PID') and the kind enum, leaving browser, channel, headless, timeoutMs, executable, windowTitle, args and cwd entirely unexplained.

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?

States specific verbs (launch, attach) and concrete resources (web/Electron/desktop browser, CDP endpoint, native PID), so the operation is unambiguous. It doesn't differentiate itself from siblings like session_list/session_close, but the core purpose is clear.

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?

No guidance on when to use this versus other session tools, no prerequisites, no mention that a session must exist before ui_* or test_run tools are used. Usage is only implied by the name 'session_start'.

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