Spawn subagent

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

With no annotations, the description carries the full burden and does well: it discloses that the child starts with no parent context, that the call blocks by default until finish or approval, that it returns a resumable agent_id plus name and timeout_seconds, that a full concurrency cap waits rather than fails, and that model is mandatory unless the agent config pins one. Gaps remain around the approval handoff (which sibling resolves it) and what happens to worktree state after completion.

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 and blocking behavior are front-loaded, and each sentence carries operational detail. The closing advisory-output paragraph is valuable policy but sits apart from the invocation mechanics, making the block slightly longer than strictly necessary.

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 15-parameter tool with a nested permission object and no output schema, the description covers returns, blocking/concurrency, timeouts, and model requirements well. It leaves group_id, max_turns, depends_on ordering semantics, result_schema validation, and permission handling to the schema, which is reasonable given high coverage.

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 87%, so baseline is 3, but the description adds real semantics: the timeout cap defaults to the agent config and is echoed as timeout_seconds, name is auto-assigned when omitted, and model has no default at any layer. It also flags that subagent output should be treated as advisory. Minor redundancy with schema text keeps it below 5.

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 a specific verb (delegate/spawn) and resource (fresh subagent), enumerates the supported backends, and explicitly frames itself as mirroring the native Agent tool. An agent can distinguish this from stop_agent, send_message, or get_agent_result without opening the 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?

Gives a clear when-to-use rule (implementation work needing search, iteration, or verification should go to a subagent rather than inline) and describes the default blocking behavior versus run_in_background. It does not name sibling tools (wait_for_agents, get_agent_result, resolve_approval) as explicit alternatives, which keeps it from a 5.

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