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

With no annotations at all, the description bears full behavioral burden and does not carry it: it never says whether duplicates are deduplicated, whether an assert on an existing fact overwrites or errors, whether it is idempotent, or what provenance means operationally. Only the return value is disclosed.

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 short sentences, front-loaded with the action and ending with the return value; nothing is padded. The terseness is efficient, though it reflects under-specification rather than disciplined editing.

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 covers the return value, but for a mutation tool with three 0%-documented parameters and zero annotations, the description omits too much: write behavior, idempotency, namespace semantics, and relationship to the correct/retract siblings are all 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 coverage is 0%, so all three parameters (fact, source, namespace) are undocumented in the schema. The description hints that "provenance" maps to source, but never explains the fact format or what namespace scopes or defaults to, leaving the caller to guess.

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?

"Store a fact with provenance" gives a specific verb and resource, and "Returns the fact id" sets expectations. It implicitly separates the tool from memory_search and memory_retract, but never names a sibling or states the write semantics explicitly, so it stops short of full differentiation.

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?

There is no guidance on when to assert versus when to use memory_correct or memory_retract, and no mention of preconditions or ordering (e.g. search first to avoid duplicates). The agent must infer all routing from the tool names.

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?

No annotations are provided, so the description carries the full burden. It implies a report (non-destructive detection) rather than mutation, which is useful, but it says nothing about whether detected junk is altered, what permissions are needed, scope of cost/latency, or whether results are paginated.

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?

It is short, but this is under-specification rather than conciseness: a single noun-phrase fragment with no front-loaded statement of action, scope, or result. Nothing is wasted, but very little is delivered.

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 explained, but for a no-annotation tool the description should cover the namespace scope and the read-only/non-destructive nature of the audit. Both are absent, leaving the definition too thin to call confidently.

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 0% for the single 'namespace' parameter, and the description never mentions namespace at all. Its only clue is the default 'default', so the agent gets no meaning, format, or scoping guidance for the one input it must supply.

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 fragment 'Junk report: duplicates, contradictions, orphans, stale facts' tells an agent this surfaces memory quality defects, which distinguishes it from the mutating siblings memory_correct/memory_retract and from memory_search. But there is no verb and no explicit statement that this is an audit/diagnostic read, so the purpose is inferred rather than stated.

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?

No when-to-use guidance, no mention of alternatives, and no prerequisites. The agent must guess whether this runs before or after corrections, and nothing routes it against memory_search or the correction tools.

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?

No annotations are supplied, so the description carries the full burden. It never states whether the correction is destructive, irreversible, requires confirmation, or what happens to conflicting existing memories; the 'forget' example only weakly implies removal.

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?

A single short sentence with the core purpose front-loaded and no filler. It is efficient, though the fragmentary phrasing provides no structure beyond the one example.

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, but for a mutation tool with no annotations the definition is thin: no safety/irreversibility notes, no sibling routing, and an unexplained namespace default. More is needed 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 description coverage is 0%, so the description must compensate. The example ('forget that I prefer tea') usefully illustrates the expected natural-language form of the required instruction parameter, but the namespace parameter is left completely unexplained in both schema and description.

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 states a recognizable verb+resource ('correction' of memory) and gives a concrete example, so the agent knows roughly what it does. However, it does not differentiate from the close sibling memory_retract, which plausibly also removes/alters memory, leaving the boundary ambiguous.

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?

There is no statement of when to use this tool versus memory_retract, memory_assert, or memory_search. The example hints at a removal-style correction but no exclusions or selection criteria are given.

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 burden, and it does disclose one genuinely important trait: this is a soft delete ('never hard-deleted'), which tells the agent the record is retained and searchable state may change rather than data vanishing. However, it says nothing about permissions, reversibility/undo, or effects on existing references, leaving substantial behavioral gaps.

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, no filler, with the core operation stated first and the retention guarantee second. Nothing wastes the agent's attention.

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 explained. Still, for a mutation tool with zero annotations and no parameter documentation, the description is thin — it does not say what a tombstoned fact does to future searches or whether the retraction can be reversed, both of which materially affect how an agent uses 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 0%, so the description would need to compensate for all three undocumented parameters. It does not: 'reason' and 'namespace' carry no explanation of format, expected values, or whether reason is required for audit purposes, and only fact_id is self-evident from its name.

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?

'Tombstone a fact' gives a specific verb and resource, and 'tombstone' is a precise technical term that distinguishes this from memory_search or memory_assert. It does not explicitly contrast itself with the closest sibling, memory_correct, so an agent must infer the boundary.

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?

The description offers no when-to-use guidance and never mentions the alternatives (memory_correct, memory_assert) that an agent would weigh against it. A reader cannot tell from this text whether retraction is preferable to correction in a given situation.

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 burden, but it does disclose one meaningful behavioral guarantee: tombstoned (retracted) facts never surface, which tells the agent results always reflect live truth. Beyond that it says nothing about ranking/similarity semantics, how k interacts with scoring, or namespace isolation, so it only partially covers the behavior an agent needs.

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?

A single tight sentence with the scope statement front-loaded and no filler. It is efficiently written, though the terseness contributes to the under-specification noted elsewhere rather than being a model of completeness.

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-value explanation is not required, but for a three-parameter search tool with zero annotation coverage and zero parameter descriptions the definition is too thin. An agent cannot tell what k or namespace do or how results are ranked from this description alone.

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 0% and the description adds nothing about any of the three parameters. The role of k (top-k result count, default 5) and namespace (isolation scope, default 'default') is left entirely unexplained, and the description does not compensate for the schema gap.

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 (Search) and the resource/scope (the current-truth view), and the parenthetical clarifies that retracted facts are excluded. It does not explicitly contrast itself with siblings like memory_audit or memory_assert, but the read-vs-mutate distinction is apparent from the name.

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?

No explicit when-to-use guidance, no conditions for choosing this over memory_audit, and no stated prerequisites (auth, namespace setup). The parenthetical about tombstoned facts is a behavioral fact, not routing guidance, so an agent must infer that this is the default read path.

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