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

Annotations already declare destructiveHint=true and idempotentHint=true, so the safety profile is partly covered. The description adds real value beyond that: it discloses that this is a full replacement of the published graph (not a patch), that it never touches Git/GitHub or inspects remotes, and that changeNarrative must be authored manually rather than generated. Missing is what happens on a version conflict or whether the prior graph is recoverable.

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?

Two sentences, verb-first and front-loaded with the core action, followed by exclusions and the narrative requirement. Every clause earns its place; there is no filler or restatement of the name.

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 an 8-parameter, nested-object mutation with no output schema, the description covers the essential context an agent needs: identity resolution rules, the replace-vs-patch distinction, manual narrative authoring, and the fact that no remote/Git side effects occur. It leaves the node/edge payload semantics and the payload-size expectations to the schema, which is reasonable, but a note on the expectedProjectVersion conflict outcome would have completed it.

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 roughly half the parameters (nodes, edges, sessionId, mappedRevision) carry no semantic help in either place. The description does add meaning for changeNarrative by stating it must contain four manually written plain-language sections, and it clarifies what projectId must be (an owned ID, not a path or URL). It does not compensate for the undocumented nested node/edge structures.

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 description opens with a precise verb+resource: 'Replace the privacy-filtered published graph for exactly the existing Brain Scanner project identified by projectId.' It further separates itself from neighbors by negation ('Never creates a project, infers one by display name, inspects a repository remotely, or changes Git'), which rules out create_project/map_project-style behavior. It stops short of naming the sibling it contrasts with (e.g., publish_project_graph or patch_project_graph), so a 5 is not warranted.

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?

Usage context is implied rather than stated: the caller learns it targets an already-existing linked project and that a publishing agent must supply narratives, but there is no explicit 'use this instead of patch_project_graph when…' or 'do not call when the graph does not exist yet.' The scope exclusions gesture at when-not, but the relationship to sibling mutation tools is left to inference.

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