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

Beyond the annotations, it discloses a genuine refusal condition ("Refused if you do not own that agent (or it was retired)") and what the call yields, which annotations alone do not convey. However, the central sentences about `before_every_action`, `live: false`, `until_live`, `blocked: true`, and `until_unblocked` describe no parameter of this tool, so the behavioral disclosure is partly about an unclear external rule structure rather than this call.

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 lead sentence is well front-loaded with the verb, resource, and return value. The remainder is a dense run-on mixing ownership rules, downstream routing, and pseudo-parameter flags, which dilutes the signal for a two-parameter tool.

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 output schema, the description does supply the key return information (`run` handle and `config_revision`) and the ownership refusal condition. It omits important behavior for a non-idempotent call – notably that calling twice likely creates two separate runs – and spends space on non-existent parameters instead.

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 parameter burden, and it does explain `agent_key`'s origin and defines `started_by` values (person vs. schedule origin). That compensation is offset by noise: it explains several flags (`live`, `blocked`, `before_every_action`) that are not inputs here, which can mislead an agent into hunting for parameters that do not exist.

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 first sentence gives a specific verb and resource ("Begins a run as the agent you name") and explicitly sources `agent_key` from the sibling list_my_agents, which distinguishes it from register_my_agent and report_agent_run. It also states what is produced (a `run` handle and `config_revision`), but the purpose is partially buried under dense downstream jargon.

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 concrete downstream usage instructions ("Pass `run` to check_premises and verify_premises") and points at list_my_agents for the required key, so the caller knows how to obtain inputs and what to do next. It does not say when to prefer this over other session-starting siblings or when a caller should skip starting a run entirely.

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