tandem_work

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 the plan invariant (one final integration step depending transitively on all others), the operation_id retry-vs-new-request semantics, the need to re-read the card for the current revision before mutating, and that autonomous grants/reconciliation are out-of-band CLI actions. It stops short of stating what happens on a stale revision or permission failures.

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 is front-loaded in the first sentence, followed by short, information-dense sentences grouped by concern (create, later mutations, idempotency, exclusions). No filler, though the telegraphic style occasionally makes the action-to-parameter mapping hard to follow.

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 tool is highly complex (18 actions, nested plan object), and an output schema exists so return values need not be described. The description covers the creation contract, mutation preconditions, and idempotency, but omits the ordered propose/agree/claim/submit/accept lifecycle and stale-revision conflict behavior, which an agent would need for non-create actions.

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 50%, so the description must compensate: it covers action, expected_revision, operation_id, plan, and work_id (create path) meaningfully. But many parameters (condition, blocker_id, submission_id, resolution, commit, note, evidence) are left entirely to the schema, and the description does not explain which of the 18 actions require which of them.

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 ('Maintain') and resource ('one durable shared task and its role-bound checklist'), which tells an agent this is the lifecycle manager rather than a read-only or artifact tool. It does not explicitly distinguish itself from siblings like tandem_review or tandem_wait, so it stops short of a 5.

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 clear when-rules: use action="create" with expected_revision=0 and omit work_id; for later mutations fetch the card first and use its current revision; list/get/history need no operation_id. It even cautions that 'waiting observes committed changes, not permission to start an agent.' However, it never says when to reach for tandem_work versus the other tandem_* siblings.

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