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

With no annotations, the description carries the full behavioral burden. It does disclose a destructive effect ("retire those memories") and that recall duplicates this behavior implicitly, which is valuable. But it never defines what retirement means (deletion vs. flagging vs. mark_outdated), nor whether the check is report-only, leaving real ambiguity about consequences.

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 action and its scope front-loaded before the sibling cross-reference. 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?

An output schema exists, so return values need not be explained, and the description covers purpose, triggering condition, and the sibling overlap. The only shortfall is the undefined semantics of "retire," which matters for a mutating, no-parameter tool.

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?

The tool takes zero parameters, so there is no parameter meaning to convey; the description correctly explains the implicit scope (all memories with linked files) instead of inventing arguments. Baseline 4 applies for a parameterless tool.

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: check whether a memory's linked files changed since save, then retire those memories. It also names the sibling that performs the same work (recall), so an agent can separate it from the other memory tools.

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 an explicit condition for calling it ("directly after big changes") and names the alternative (recall does this automatically). The phrasing is slightly ambiguous about why to call it at all if recall is automatic, but the when-to-use signal is present and concrete.

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. It usefully discloses a side effect beyond the name — it re-links the memory to the current version of its files — but says nothing about required permissions, reversibility, or what happens to conflicting/stale links.

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, front-loaded with the core action and followed by the qualifying scenario and the side effect. No filler, though the parenthetical adds a level of nesting that slightly slows reading.

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?

Output schema exists, so return values need not be described, and the tool is a simple single-parameter mutation. However, with zero annotations the description should cover more of the mutation's behavior — permissions, reversibility, or how re-linking interacts with check_stale/sync_to_files — to be fully complete.

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?

One required parameter with 0% schema description coverage, and the description never mentions it. The parameter is a single self-evident identifier ('memory_id') for the memory being confirmed, so the gap is small, but no format or sourcing guidance is added.

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 ('Mark a memory as still true') and clarifies the state transition being reversed (retired when a linked file changed). This implicitly contrasts with mark_outdated and replace_memory, though no sibling is named explicitly.

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 parenthetical supplies a concrete trigger condition — use it after a memory was retired because a linked file changed but the fact itself didn't. It does not state the inverse case (when the fact is no longer true, use mark_outdated) or name alternatives, so it stops short of a full when/when-not routing guide.

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. The naming ('explain', 'show') strongly implies a read-only inspection, and the description discloses the two categories of information returned (origin and subsequent history), but it never states read-only semantics, whether any state changes, or permission 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?

Two sentences, front-loaded with the core purpose and closing with the usage cue. No filler or restated name/title.

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 structure need not be explained, and the description already says what is surfaced. For a one-parameter read tool this is nearly complete; only the read-only guarantee and memory_id format 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 0% for the single memory_id parameter, so the description should compensate and does not — it says 'a memory' with no hint about ID format or how to obtain one. The parameter is self-describing enough ('memory_id', title 'Memory Id') that the gap is mild rather than fatal.

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: 'show where a memory came from' plus 'everything that has happened to it since'. This clearly separates it from siblings like recall (retrieval) or check_stale (staleness), though it never names an alternative explicitly.

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 triggering condition: 'Use it when a memory looks wrong or surprising.' That is real when-to-use guidance, but there is no when-not guidance and no named alternative for related diagnostic tasks such as list_conflicts or check_stale.

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. 'List' implies a read-only operation, which is a reasonable signal, but the description does not confirm non-mutation, disclose whether results are ranked or filtered, or say anything about the decision workflow the pairs feed into.

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?

One sentence, front-loaded with the verb, and every clause adds scope or intent. No filler or restated title.

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 structure need not be explained. The description covers scope and intent adequately for a zero-param lister, though it could have named resolve_conflict to complete the workflow picture.

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?

The tool takes zero parameters, so there is nothing for the description to disambiguate; baseline 4 applies. The description correctly signals that scoping is implicit (current project) rather than parameterized.

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 (List) and resource (pairs of memories that may contradict each other), with the scope confined to 'this project'. It implicitly distinguishes itself from resolve_conflict by framing the output as items that 'need a decision', though it does not name that sibling.

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 phrase 'need a decision' hints that this is a discovery step preceding resolve_conflict, but the description never states when to call this versus resolve_conflict or check_stale. Usage is implied rather than spelled out.

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 does disclose the important behavior: the memory disappears from recall while its history and the retirement reason remain visible. That is exactly the state semantics an agent needs for a mutation. It does not cover reversibility (can a mark be undone?), permissions, or error behavior, which keeps it from a 5.

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 tight sentences, front-loaded with the action and followed by the consequence. Every clause carries information (action, scope, recall effect, history retention, rationale for the retained reason) with no padding.

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, and the description covers the core state transition and what is preserved. The main remaining gap is the undocumented parameters and unstated failure/undo behavior for a mutation tool with no annotations.

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% — neither memory_id nor reason has any description. The description partially compensates by framing the target as "a memory" and characterizing reason as "a short reason," giving a hint about length and intent. It adds no format, ID convention, or constraint detail for memory_id, so the gap is only partly closed.

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+target state: "Mark a memory as no longer true." The effect clause ("stops showing up in recall but keeps its history") cleanly separates it from a hard delete or from replace_memory, which is useful for an agent scanning siblings. It stops short of naming an alternative tool explicitly, so it is clear but not fully differentiated.

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 is implied by the state change described (memory is no longer true), but there is no explicit guidance on when to prefer this over replace_memory (correct the content) or confirm_memory (re-affirm), and no statement of preconditions such as the memory needing to exist. An agent can infer the case but is not routed.

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 burden and it does disclose a real behavioral trait: only active memories are returned and outdated/replaced ones are skipped. It does not mention auth requirements, rate limits, or result ordering, but the filtering semantics are a meaningful addition.

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 short sentences, front-loaded with purpose followed by the call trigger and the result-filtering caveat. Nothing is redundant or padded.

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 structure need not be explained. The description covers purpose, timing, and result scope for a read-only search tool, but leaves the limit parameter semantics and any guidance on query phrasing unstated.

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 the description must compensate. 'By meaning' usefully defines the query parameter as a semantic/natural-language search rather than keyword matching, but the limit parameter is never explained and no format or defaults are given beyond the schema.

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 ('Search this project's shared memory') plus the retrieval mechanism ('by meaning'), which separates it at a glance from siblings like remember, mark_outdated, or list_conflicts. An agent can pick this tool without opening the 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?

Gives explicit timing guidance: 'Call it at the start of every task, and whenever project-specific knowledge (setup, conventions, past decisions) might matter.' That is strong when-to-use context, though no alternatives or exclusions are named (e.g., when to prefer explain_memory or check_stale).

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 must carry the behavioral load. It does disclose real traits: the fact persists across sessions and tools, and dependent files cause the memory to be flagged when those files change. However, it says nothing about what happens when a conflicting or duplicate fact already exists (notable given siblings like list_conflicts and resolve_conflict), nor about failures or permissions.

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 purpose, then usage, then content guidance, then the auxiliary parameter. Every line contributes, though the 'files' paragraph reads like schema documentation appended with a leading lowercase parameter name, which slightly breaks the prose flow.

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, and the description covers persistence semantics, trigger conditions, and the file-linkage behavior well. The remaining gap is conflict/duplicate handling, which is meaningful in a toolset that exposes list_conflicts, resolve_conflict, and mark_outdated.

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 the description must compensate, and it largely does. The 'files' parameter is fully explained (repo-relative paths, concrete example, and the purpose of the linkage), and 'fact' receives content guidance via the good/bad fact examples, which is more useful than a bare type declaration.

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 specific verb and resource ('Save a durable fact about this project') plus the scope that matters: durable across future sessions and any AI tool. It clearly separates itself from an external alternative (built-in memory or notes file), but it never distinguishes itself from its actual siblings such as recall, replace_memory, or mark_outdated, so an agent gets no routing help within the tool family.

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?

It gives an explicit when-to-use trigger ('whenever the user asks you to remember something about the project'), an explicit when-not ('temporary task details or anything in the code already'), and contrasts with a competing mechanism ('use this, not a built-in memory or notes file'). Good-fact and bad-fact examples remove remaining ambiguity.

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 does well: it discloses that the old memory disappears from recall but retains its history, which is the key non-destructive/retirement behavior an agent needs. It still omits whether the operation is reversible, what happens if new_id doesn't exist, and permission requirements.

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 tight sentences, front-loaded with the core action, and the parenthetical example earns its place by disambiguating the semantics. No filler.

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 no explanation, and the behavioral effect is covered. However, for a mutation-style tool with zero annotations and two undocumented required ID parameters, the description leaves an agent guessing about ID provenance and irreversibility.

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% and both required parameters are undocumented titles ('Old Id', 'New Id'). The description implies the old/new relationship but never states that these are memory identifiers, nor any format or validity constraints, so it only partially compensates for the coverage 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 and resource ('Retire an older memory') plus the causal condition that defines it ('because a newer one replaces it'), with a concrete example. This distinguishes it from siblings like mark_outdated, which retires a memory without a superseding replacement.

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 selection condition is clear and concrete: use this when a newer memory supersedes an older one, and the Postgres/MySQL example makes the scenario unambiguous. It does not explicitly name mark_outdated as the alternative for retiring without a replacement, nor state exclusions, so it stops short of a 5.

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. It usefully discloses the destructive outcome of replaces_old=True ('the old one is retired') and the non-destructive alternative, but says nothing about reversibility, permissions, error behavior, or whether other memories referencing the retired one are affected.

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?

Tight and front-loaded: the core action comes first, then the two flag branches, then the escalation rule. No filler sentences, though the value-branch phrasing is slightly terse.

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, and the acceptance/rejection semantics of the boolean are covered. The main omission is any prerequisite about how a conflict is identified before calling this tool.

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 the description must compensate. It fully explains the pivotal boolean replaces_old with both branches, which is the hardest parameter; however, new_id and old_id are left entirely undocumented and their roles are only inferable from their names.

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 ('settle') and resource ('a possible conflict'), and the flag semantics make the outcome unambiguous. It does not explicitly differentiate itself from siblings like list_conflicts or replace_memory, so the 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?

'Ask the user if you're not sure' gives a useful uncertainty rule, and the tool is implied to be used when a conflict is detected. However, it never names which sibling to use instead (e.g. list_conflicts to find conflicts, replace_memory for a straight replacement), leaving the routing decision to inference.

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 does well: it discloses the write target, that only the 'managed section' is touched (implying other file content is preserved), and the observable downstream effects (shared with teammates and non-Aether tools, visible in code review). It leaves one real ambiguity — whether the tool itself commits or merely writes files that are later committed — and says nothing about idempotency or overwrite behavior.

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, front-loaded with the action and target, followed by the rationale and the trigger condition. No filler or restated name/title; every clause adds either scope, justification, or the call condition.

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 no explanation, and with zero parameters the remaining burden is behavioral. The description covers destination, audience, and trigger, but does not resolve whether the tool performs the commit or whether repeated calls are safe, which are the two things an agent would still want before invoking 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?

The tool takes zero parameters, so there is nothing for the description to clarify beyond the schema. The baseline for a no-argument tool applies, and the description correctly spends no words on parameter explanation.

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 specific verb and destination ('Write this project's current memories into the managed section of AGENTS.md / CLAUDE.md'), which is materially different from siblings like remember (writes to the cache) or recall (reads). An agent can distinguish this persistence/sync action from the rest of the family 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?

It gives explicit trigger conditions: 'Call it when the user asks, or after a batch of new memories.' That is clear when-to-use guidance, though it names no alternative or exclusion (e.g. it does not say what to use instead when only a single memory changed).

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