Lians Agent Memory

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

With no annotations, the description carries the transparency burden. It discloses the scanning mechanism, the two flag types (FUTURE_EVENT, LATE_REVISION), and the is_clean output. This goes beyond a basic summary, though it doesn't state read-only behavior or error conditions.

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, each earning its place: purpose, mechanism, output/significance. No fluff, front-loaded with the core function.

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 simplicity (1 parameter, no output schema), the description covers the essential aspects: what it checks, what flags it returns, and what a clean result signifies. Missing edge cases or error handling, but these are less critical for 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 100% for the single parameter. The description only references 'given simulation date' without adding detail beyond the schema's ISO-8601 UTC timestamp definition, so it meets the baseline but adds no extra 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?

The description clearly states the tool's purpose with a specific verb ('Detect lookahead bias') and resource ('a backtest simulation'). It distinguishes itself from memory retrieval siblings by focusing on validation and analysis of memory store contents.

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 context is clear: it's used to produce a clean report for risk committees. It implies when to use it (when validating a backtest) but doesn't explicitly list exclusions or alternatives. Given the specialized nature, this is sufficient.

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 discloses key behaviors: ordering by event_time, inclusion of superseded versions, and automatic entity normalization (Apple Inc./ISIN/AAPL). It omits limit semantics and error behavior, but the disclosed traits are substantial.

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 with no filler: the first states the core action, the second gives a use case, and the third adds behavioral details. Information is front-loaded and every sentence 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?

Given no output schema, the description covers the main expectations: what is returned (versions), ordering, superseded handling, and normalization. It does not explicitly describe the response structure or pagination, but for a tool of this complexity it is reasonably 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?

Schema descriptions cover ticker and metric, and the description adds entity normalization and time ordering context that clarifies parameter usage. However, the 'limit' parameter remains undocumented in both schema and description, preventing a higher 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 uses a specific verb ('Return') plus resource ('every recorded version of a structured fact') and distinct behavior ('ordered by event_time'), clearly distinguishing it from sibling tools like recall or remember. It leaves no ambiguity about what the tool does.

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 clear context with a concrete example ('show me how AAPL EPS evolved over the last four quarters') indicating it is ideal for time-series views. It does not explicitly name alternatives or when not to use it, but the context is strong enough for an agent to select it appropriately.

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. It discloses key behavioral traits: returns up to 20 conflicts, includes both memory contents, and is scoped to open conflicts by default. It does not mention side effects or auth, but as a read-only listing tool, this is reasonably transparent.

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 two sentences, front-loaded with the core action, and every clause provides useful information (definition, use case, return limit, contents). No redundancy or 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?

For a simple one-parameter tool without annotations or output schema, the description covers the essential aspects: what the tool does, why to use it, and what it returns. It briefly misses explaining the full range of statuses available via the schema, but this is a minor gap given the schema already enumerates them.

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 no meaning to the 'status' parameter beyond echoing 'open' as the default. It does not explain other enum values (accept_a, accept_b, dismissed) or how filtering works, leaving the parameter semantics to the schema alone.

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 lists conflict flags and defines what constitutes a conflict (two sources reporting different values for the same fact at the same event_time). This specific verb+resource+definition distinguishes it from sibling tools like recall or memory_lineage, which serve different memory-related purposes.

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 context for when to use: 'Use this to surface data quality issues before they affect decisions.' It does not explicitly name alternatives or exclusions, but the purpose is clear and distinct from siblings, making the intended use evident.

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?

The description discloses a key behavioral trait: negative feedback flags evidence for review and never silently deletes it. Since no annotations are provided, this adds important context about side effects and safety, though it does not mention what happens with positive feedback or other signals.

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 a single, front-loaded sentence with no redundancy. It efficiently captures the tool's purpose and a critical behavioral constraint.

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 feedback-recording tool, the description covers the core purpose and a key behavioral consequence (no silent deletion). No output schema exists, but the tool's return value is not essential for the agent to select and invoke it correctly. Still, a mention of what happens after recording (e.g., confirmation) would make it more 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?

Schema description coverage is 0%, so the description must compensate. It does enumerate the signal enum values, but it fails to explain the meaning of memory_id, note, and outcome. These parameters remain ambiguous, leaving gaps beyond what the schema 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 uses the specific verb 'Record whether' and clearly lists the signal values (helpful, incorrect, outdated, duplicate, ignored), making it distinct from sibling tools like recall and remember. The resource (recalled memory) is explicit, so purpose is unambiguous.

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 implies the tool is used after memory recall ('Record whether a recalled memory...') but does not explicitly state when to use it over alternatives like backtest_check or list_conflicts. No exclusions are provided, leaving usage guidance implied rather than explicit.

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 burden. It indicates a read-only style operation by using 'Return', and discloses what is returned (full history, every prior version, chain of updates). However, it doesn't mention limitations, error behavior, or any side effects, so transparency is partial.

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 consists of two concise, front-loaded sentences. The first states the action and object, the second provides use cases. No redundant or vague 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?

For a simple one-parameter tool with no output schema, the description covers purpose and usage adequately. It explains what is returned and when to use it, though it omits edge-case behavior like invalid IDs or ordering. Given the low complexity, this is sufficiently 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?

Schema coverage is 100%: the parameter memory_id is described in the schema as 'UUID of any memory in the lineage chain'. The tool description adds no further parameter semantics beyond the schema, so baseline 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 specifies the verb 'Return' and the resource 'supersession history of a memory', with additional detail about every prior version and the chain of updates. This distinguishes it from sibling tools like recall or fact_history by focusing on lineage and supersession.

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?

Explicitly states when to use: 'how did this guidance number evolve over time?' or 'investigating why a memory was replaced'. This gives clear context, though it does not mention when-not-to-use or alternative tools explicitly.

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 of behavioral disclosure. It clearly explains that superseded facts are excluded at the DB layer and only currently-valid memories are returned, which is valuable behavior beyond the basic 'retrieve' action. It does not cover auth or rate limits, but for a read-only-like tool the core behavior is well 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?

The description is two sentences plus an example, entirely front-loaded and free of fluff. Every clause adds meaning: what it retrieves, the current-validity constraint, when to call it, and how to narrow results.

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 is a retrieval operation with no output schema and simple parameters, the description covers most key aspects: purpose, behavior, usage timing, and a filter example. Missing details are k's effect on result count and return format, but for a tool of this simplicity it is reasonably 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?

Schema description coverage is only 33% (only filters has a description). The description adds a useful example for filters ({ticker: NVDA}), but does not explain k or query beyond the schema. Since k has a default and query is self-explanatory, the compensation is moderate but incomplete.

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's purpose: 'Retrieve the most relevant CURRENT memories for a query.' It uses a specific verb and resource, and the emphasis on 'CURRENT' distinguishes it from siblings like recall_at or memory_lineage, which deal with historical or lineage data.

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 gives explicit when-to-use guidance: 'Call this before answering any question that may be in memory.' It also provides a concrete filter example. However, it does not explicitly mention when not to use it or directly contrast with alternatives like recall_at or reconstruct.

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 burden and does disclose the key behavioral trait: later superseding updates are excluded for true point-in-time recall. It does not explicitly state read-only safety, but 'Retrieve' implies a read operation, and the temporal exclusion is the most critical behavioral aspect.

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 three sentences and front-loads the core purpose. However, the final sentence comparing mem0 and Zep is somewhat promotional and not necessary for using the tool, adding unnecessary length.

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 read-oriented retrieval tool, the description adequately covers the essential temporal semantics and provides a use case. It does not describe return format or pagination, but given the straightforward nature of a memory query, this is acceptable. Overall, it provides sufficient context for correct usage.

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 only 33% (only as_of_iso has a schema description). The description adds no meaningful detail about 'query' or 'k', and merely reinforces the purpose of as_of_iso without providing syntax or format guidance. It fails to compensate for the low schema coverage.

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 retrieves memories valid at a specific past point in time, with a concrete compliance/audit example. It distinguishes itself via 'true point-in-time recall' and notes that later superseding updates are excluded, making it specific and differentiated from regular recall.

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?

Explicitly states 'Use for compliance and audit' with an example query, and explains that later superseding updates are excluded, which helps select this tool over others. However, it does not mention alternative sibling tools or when not to use it.

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 carries the full burden. It details the return payload (memory state plus timestamped, hashed event log) but does not disclose whether the operation is read-only, requires specific permissions, or has performance implications.

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 two sentences, leading with the main purpose and then giving use cases. There is no filler and it is well front-loaded.

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 no output schema and no annotations, the description provides a high-level overview but lacks detail on the exact output structure, pagination, error handling, or edge cases such as future dates or missing data. For a complex reconstruction operation, more context would be expected.

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 schema covers only 50% of parameters (query has a description, as_of_iso does not). The description clarifies that as_of_iso specifies the point in time for which memories were valid ('valid at as_of'), partially compensating for the schema gap, but it does not add meaning to the query parameter 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 uses a specific verb 'reconstruct' and clearly identifies the scope as 'complete memory state and full audit event trail at a past point in time'. It differentiates from siblings like recall_at by emphasizing the inclusion of the hashed event log.

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 explicitly states use cases: 'Use for regulatory audit submissions and trade reconstruction.' This provides clear context, but it does not name alternative tools or state when not to use, so it lacks explicit 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 are provided, so the description carries full burden. It does disclose a key behavior—metadata enables automatic replacement of stale guidance numbers—but it doesn't mention return values, side effects, or any destructive potential. Adequate but not comprehensive.

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 two sentences, front-loaded with purpose, then immediately gives critical usage guidance. No fluff—every sentence 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 write tool with 4 params and no output schema, the description covers the essential purpose and parameter semantics sufficiently. It lacks return value details and explicit side-effect disclosure, but is still reasonably complete for typical use.

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 75%, and the description adds meaningful semantics beyond schemas: it explains the event_time_iso field must reflect when the event occurred, not now, and shows how metadata supports supersession. This enriches parameter understanding.

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 a specific action ('Store a financial fact, observation, or decision in persistent memory') with a clear resource (persistent memory). The verb 'store' distinguishes it from sibling tools like 'recall' and 'reconstruct', which are read operations.

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 provides strong guidance on how to use the tool (e.g., event_time_iso semantics, metadata for supersession) but does not explicitly state when to use it vs. alternatives, nor any exclusions or when not to use it. Usage context is implied but not explicit.

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