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

With no annotations, the description must carry the behavioral burden. It does disclose a meaningful trait — the fact lands in a *shared* memory readable by all connected Claude windows — but omits whether saves append or overwrite, whether there are size/count limits, and how the entry is later identified for deletion.

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?

Three sentences, front-loaded with what/where, then when, then how. Little waste; 'Write it clearly and specifically' is mild instruction fluff but still earns its place as content guidance.

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 simple two-parameter write tool with no output schema and no annotations, the description covers purpose, trigger, granularity and visibility well enough to call it correctly. Only the append-vs-overwrite and post-save identification details are missing.

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 100%, so both parameters are already documented in the schema and the baseline applies. The description adds only the granularity hint ('one important fact') and encourages specificity in the content, which is marginal beyond structured data.

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+resource ('Save one important fact ... to the shared memory') and bounds scope with examples (preferences, projects, decisions), plus the cross-window visibility detail. It does not explicitly name or distinguish itself from siblings like memory_search/memory_delete, but the write verb makes the distinction inferable.

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 a concrete trigger: 'Use when the user says "remember this" or states something worth keeping.' No exclusions or when-not guidance are offered, and it never routes the agent to memory_search for deduplication, so it stops short of the top band.

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

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

With no annotations, the description carries the full behavioral burden and does disclose the critical side effect: unread messages are 'marked read' as a result of the call. It omits permissions, concurrency, or rate-limit details, but the mutation disclosure is the most important trait and it is present.

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?

Three tight sentences with zero filler, leading with purpose and side effect before usage timing. Every clause earns its place.

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 small read tool with no output schema and no annotations, the description covers purpose, default behavior, the state-changing side effect, and call timing. Return format is unspecified but not essential for correct invocation.

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 baseline would be 3, but the description adds real meaning: it explains the default read-unread behavior and clarifies that all_history=true 'lists the whole exchange,' which sharpens that parameter beyond the schema's wording. Only the topic parameter is left without added semantics.

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: 'Read messages addressed to you (and broadcasts).' This clearly distinguishes it from write-oriented siblings like msg_send, though it never names msg_board or explains how it differs from that read-like sibling, keeping it 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 concrete timing guidance ('Call before starting work and after finishing a chunk') plus the key default-vs-history mode. It does not name an alternative sibling or an exclusion condition, so it is clear context without full routing.

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

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

With no annotations to lean on, the description still discloses the non-obvious behavioral trait that governs this tool: delivery is pull-only, so the recipient will not be notified. It omits return-value behavior, failure modes (e.g. unknown recipient), and whether messages can be edited or recalled.

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?

Three tight sentences with the purpose front-loaded and delivery semantics before addressing details. The final sentence largely repeats the schema's own "to" description, costing a little efficiency.

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 simple messaging tool with no annotations and no output schema, the description covers purpose, delivery model, and addressing adequately. It leaves minor gaps around return value and recipient-existence behavior that an agent might want.

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 100%, so all four parameters are already documented, including the "to" = name or "all" convention. The description merely restates the "to" semantics and adds no syntax, format, or constraint beyond the schema, so the baseline 3 applies.

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+resource ('Leave a message for another Claude window (shared mailbox)') and immediately distinguishes itself from the pull-based siblings by noting the recipient only sees it via msg_read. An agent can tell this apart from msg_read/msg_board without opening any schema.

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?

Clearly frames the asynchronous delivery model ('nothing is pushed') and explains the addressing convention, which tells the agent when this communication channel is appropriate. It stops short of explicit when-not guidance or naming msg_board as the alternative for browsable notes (that hint lives only in the schema).

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