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

Annotations already declare destructiveHint=true and readOnlyHint=false, but the description adds real value beyond them: it states that composition is kicked off, that composition errors are returned, that the running router can break, and that dryRun returns updatedGateway while deleting nothing. Lacks permission/auth requirements, which is the only notable gap.

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?

Front-loads the action and consequence, then dryRun behavior, then required inputs. Mostly tight, but the clause 'This deletes the subgraph from the variant and can break the running router' partly repeats the opening sentence about removal and composition.

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?

An output schema exists, so return values need not be described, yet the description usefully notes composition errors and updatedGateway for the dryRun path. Combined with the destructive annotation, an agent has enough to invoke this safely; only the three required params lack descriptive depth.

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% (only dryRun is documented in-schema), and the description's explanation of dryRun largely duplicates it. For the three undocumented required params it merely names them ('provide the graph ID, the variant name, and the subgraph name') without adding format, sourcing, or naming-convention detail an agent couldn't infer from the keys themselves.

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 (delete/remove) and resource (subgraph from a variant), plus the side effect (start composition). It explicitly frames itself as the equivalent of `rover subgraph delete`, which distinguishes it from siblings like DeleteGraph (whole graph) and PublishSubgraph (write).

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 concrete operational guidance: 'Set dryRun to true first' and explains the consequence of not doing so (can break the running router). That is clear context for invocation, though it doesn't explicitly contrast against any sibling tool to say when this is the wrong choice.

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