stagenth · AI 记忆库

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

The description discloses several important behaviors: cost per usage, automatic vector computation, deduplication with timestamp refresh, per-user limit of 2000, and auto-refund on failure. However, it does not specify what the tool returns on success (e.g., a memory ID or confirmation), which is a gap given the absence of annotations or output schema.

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?

The description is concise at around 100 words, with effective use of newlines and bullet-style formatting. Key points are front-loaded (cost, cross-session retrieval, vector computation). It could be slightly more structured (e.g., using actual bullet points) but remains efficient.

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?

The description covers input limits (4000 chars, 2000 memories), dedup, vector computation, cost, and refund. However, it omits what the tool returns after a successful operation. Given no output schema, this is a notable gap for an agent to interpret the result.

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%, with each parameter having clear descriptions (e.g., content length, tag format, project namespace, TTL range, importance scale). The tool description adds no additional semantics beyond the schema, so a baseline score of 3 is appropriate.

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 clearly states the tool saves persistent memory across sessions and clients, with cost and dedup details. However, it does not explicitly differentiate from the sibling tools like memory_update or memory_search, leaving ambiguity about when to add versus update.

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 does not provide guidance on when to use memory_add versus its siblings (memory_delete, memory_list, memory_search, memory_update). While it mentions memory_search in passing, there is no explicit statement of appropriate contexts or exclusions.

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 exist, so the description carries full burden. It discloses that the operation is free (0 credit), batch, and returns a detail list. This adds behavioral context beyond the implicit destructive nature.

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?

The description is extremely concise with two sentences, front-loading the main action and adding essential constraints and details with zero wasted text.

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?

Given the tool has only one parameter, no output schema, and four siblings, the description covers the purpose, scope, cost, and return format adequately without missing critical aspects.

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% and the parameter 'memory_ids' has a clear description in the schema. The tool description does not add any additional semantic information beyond what the schema already provides.

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 clearly states 'batch delete memories' and specifies 'only delete your own', which is a specific verb and resource. It distinguishes from siblings like memory_add and memory_list.

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 includes a clear usage constraint ('only delete your own'), but does not explicitly mention when to use alternatives or exclusions. However, it provides context about batch size and return details.

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?

描述了费用为0这一行为特性，但在缺少注释的情况下未明确说明只读性质。对列出操作而言，行为透明性足够但可改进。

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?

一句话清晰概括用途、排序、分页和费用，无冗余信息。

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?

对于简单列表工具，描述覆盖了核心功能，但未说明返回格式或排序方向。总体完整。

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?

模式覆盖度100%，描述未为参数添加新语义，仅涉及分页和总数，与模式已有信息一致。

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?

描述明确了动词'列出'、资源'记忆'、排序方式'按最近更新'、分页和返回总数，与兄弟工具（memory_add等）明显区分。

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?

未明确给出何时使用或何时不使用，但根据兄弟工具名称（memory_search等）可推断用途。缺乏明确的替代方案说明。

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 provided, so description carries full burden. It discloses the algorithm (cosine similarity + keyword weight + importance weight), score output, and automatic degradation to keyword mode when embedding service offline. Also notes the tool is free (0 credit). No contradictions.

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?

Very concise: a few sentences that front-load the purpose, followed by algorithm details and usage guidance. No wasted words.

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?

Given the complexity (4 params, no output schema, no annotations), the description covers purpose, algorithm, usage guidance, and parameters reasonably well. It lacks explicit return format details, but implies the tool returns memory items with scores. Adequate for most agents.

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 is 3. The description does not add new info beyond what's in the schema for individual parameters (query, mode, limit, project). However, it does add context about the overall algorithm and cost, which indirectly supports parameter understanding. Not enough to raise score.

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 clearly states the tool is for semantic memory retrieval by meaning, with an example ('界面配色偏好' finds '喜欢 dark mode'). It distinguishes from sibling tools (add, delete, list, update) as a search operation.

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?

Provides explicit usage context: '会话开始 / 回答涉及用户偏好、历史决策前先搜一把' (search at session start or before answering about user preferences/history). Does not mention when not to use or alternatives, but the guidance is clear and practical.

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 present, so the description must cover behavioral traits. It mentions that changing content triggers automatic recalculation of semantic vectors and is free. It also implies other fields remain unchanged ('缺省不改'). However, it does not discuss permissions, rate limits, or error cases.

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 sentence packs all essential information: fields, requirement, side effect, and cost. No wasted words.

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 5 parameters, 1 required, and no output schema, the description explains inputs and behavior adequately. It lacks details on return value and error handling, but covers the core update logic.

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%, baseline 3. The description adds value by explicitly requiring at least one updating field, explaining tags are comma-separated and replaced entirely, and noting content change triggers recalc. This goes 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?

The description clearly states the tool updates a memory and lists the updatable fields (content/tags/importance/ttl_days). It distinguishes from siblings which create, delete, list, or search memories.

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 specifies the prerequisite of providing at least one of the fields, and indicates it's free. It does not explicitly state when not to use or alternatives, but sibling tool names provide context.

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