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

Annotations already flag destructiveHint=true, but the description goes well beyond them by naming exactly what is destroyed: leads are re-evaluated against the new graph, some are moved to different steps, and pending actions can be cancelled or rescheduled. It also discloses that the graph is compiled and validated before saving, and points at flow_audit for the diff.

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-loaded with the mutation and its consequence, then the step model, then the risk profile, then the escape hatch to `reference` — a sensible order. The step-type enumeration overlaps the `type` enum already present in the schema, which is the one mildly redundant passage.

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 deeply nested, high-blast-radius mutation this covers the branching model, delay semantics, and destructive consequences, and correctly outsources field-level detail to `reference`. It does not say what a successful call returns or how validation failures surface, which matters for a compile-and-save operation with no output schema.

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 100%, so the baseline would be 3, but the description adds structural rules the schema does not encode: a TRIGGER must be the last entry in its list with children under `yes`/`no`, an omitted branch ends the flow, and `delay` is measured from the previous step. These are the semantics an agent would otherwise get wrong when nesting steps.

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 and resource ('Replace a campaign's sequence') and frames it as the authoring path ('This is how you write a campaign'), which separates it from read-side siblings like flow_get and flow_audit. It never names the closest alternatives (flow_from_template, sequence_update), so an agent still has to infer which authoring tool to pick.

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-to-use context split by campaign state: on ACTIVE campaigns it warns the call is destructive and suggests flow_audit to see what moved; on DRAFTED campaigns nothing is at stake. It also routes the agent to `reference` with topic 'sequence_steps' before guessing field shapes. It stops short of telling the agent when NOT to use it in favour of flow_from_template or sequence_update.

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