Integrates with LangChain agents to provide a verifiable memory store with exact-match recall, source citations, and provable forgetting capabilities.
Memory for AI agents that cannot hallucinate. It answers only from stored facts — with the source cited — or it honestly says "I don't know." Every guarantee below is cryptographic or true by construction, not a prompt trick.
An MCP server + Python SDK. Plug it into any agent (Claude Desktop/Code, LangChain, custom). The LLM phrases; this layer guarantees the facts.
LLMs store knowledge in weights. So they hallucinate, can't cite, can't be edited, can't forget, can't be audited. That blocks agents from any high-stakes use — legal, finance, healthcare, compliance, autonomous workflows.
Related MCP server: velesdb-memory
0% hallucination — exact match only; unknown → honest abstention.
Citations — every answer carries its source.
Provable forgetting (GDPR / right-to-be-forgotten) — the fact is really gone; signed proof; Merkle root reverts.
Valid-time — version a fact; ask "as of date T"; full history.
Merkle proofs — commit all knowledge to one hash; prove a fact's inclusion without revealing the rest.
Contradiction detection — surfaces conflicting values with both sources instead of silently picking one.
Signed receipts + determinism — tamper-evident, same query → same answer.
python3 benchmark.py)Stress-tested to 1,000,000 facts on a 7 GB CPU box, no GPU:
Metric | verifiable-memory |
Hallucination on adversarial traps | 0.0% |
Accuracy when answered / citations | 100% / 100% |
Query latency (p50 / p99) | 4.4 µs / 14 µs |
Throughput | 137,000 q/s (16 threads) |
Memory | ~1.2 GB for 1M facts (~1 KB/fact) |
Provable forget | ✅ root reverts |
vs a naive "always answer" baseline: 0% vs 100% fabrication on the same traps.
pip install verifiable-memory-mcp
verifiable-memory # MCP server over stdio
# from source:
git clone https://github.com/Mars-proj/verifiable-memory && cd verifiable-memory
python3 -m vmem.server{
"mcpServers": {
"verifiable-memory": {
"command": "verifiable-memory",
"args": [],
"env": { "VMEM_STATE": "~/.verifiable_memory" }
}
}
}Then your agent can learn_fact, recall (cited or abstains), forget (provably), prove_fact, contradictions, multihop, and more — 13 tools.
Facts are stored as data (subject, relation, object + source), indexed for O(1) exact recall; answers are exact-match-or-abstain; the knowledge state commits to a Merkle root. No vectors needed for the verifiable path → 0 fabrication by construction.
This is a memory / trust layer, not a reasoning engine and not a better chatbot. It wins on verifiability (cite-or-abstain, forget, determinism, audit), not on open-ended fluency. Pair it with your LLM: LLM = language, this = ground truth.
Need a hosted API, on-prem deployment, or help integrating verifiable memory into your agent (legal / fintech / healthcare / agent platforms)? → Pilot & enterprise: Sergey · svobodg@gmail.com
MIT licensed. PRs welcome.
All exact fact-paths between start and end entities (each path citable).
| Name | Required | Description | Default |
|---|---|---|---|
| end | Yes | ||
| start | Yes | ||
| max_depth | No |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
No annotations are provided, so the description carries full burden. It discloses that paths are 'citable', but fails to mention return behavior (e.g., empty results, pagination), authorization needs, or performance implications. This is minimal behavioral disclosure for a tool with no annotations.
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 sentence, concise and front-loaded with the core function. It wastes no words, though it could be slightly more informative without losing 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?
Given the tool has 3 parameters, no output schema, and no annotations, the description is too brief. It lacks details on return format (e.g., list of path strings or objects), error conditions, and edge cases (e.g., no paths found, excessive depth). More context is needed for a complete understanding.
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%, meaning no parameter descriptions exist in the schema. The tool description does not mention any parameters (start, end, max_depth), leaving their meaning, format, and constraints entirely unaddressed.
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 the verb 'All exact fact-paths' and the resource 'between start and end entities', making the purpose clear. It implies retrieval of multiple paths, distinguishing it from siblings like 'prove_fact' or 'verify'. However, it does not explicitly differentiate from 'multihop', which may similarly traverse paths.
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 guidance is provided on when to use this tool versus alternatives. There is no mention of when not to use it, prerequisites, or comparisons to sibling tools like 'prove_fact' or 'multihop'.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Audit knowledge for conflicts: for functional relations (one value expected, e.g. 'capital','birthdate') return any (subject,relation) holding >1 live value, showing BOTH sources — call before trusting facts that may have been updated or come from multiple sources.
| Name | Required | Description | Default |
|---|---|---|---|
| functional_relations | Yes |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
With no annotations, the description bears full responsibility. It states the tool 'audits for conflicts' and returns conflict information, implying a read-only operation. It does not explicitly say it is non-destructive, but verbs like 'audit' and 'call before trusting' reasonably convey no side effects.
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 sentence that is front-loaded with the core purpose ('Audit knowledge for conflicts') and then details the condition and output. It is reasonably concise, though slightly dense; could be split into two sentences without loss.
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 one parameter, no output schema, and no annotations, the description covers the essential: what the tool does, what input to provide, and what output to expect (pairs with sources). It does not detail output format, but the description is sufficient for an agent to understand and use the 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 schema only provides the parameter name and type (array of strings) with no description. The description adds crucial meaning: the strings should be 'functional relations' expecting one value, with examples 'capital' and 'birthdate'. This compensates well for the 0% schema description 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 audits for conflicts in functional relations, with specific examples like 'capital' and 'birthdate'. It precisely defines the output: pairs of subject and relation with multiple live values, showing both sources. This distinguishes it well from sibling tools like 'recall' or 'update_fact'.
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 advises using the tool 'before trusting facts that may have been updated or come from multiple sources', which gives clear context for when to invoke it. It does not explicitly list when not to use or mention alternatives, but the guidance is sufficient for typical use.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Permanently and PROVABLY delete a stored fact (GDPR / right-to-be-forgotten). Use when the user asks to forget or remove information — the fact is fully erased and you get a signed proof of deletion. object optional (omit to delete all values).
| Name | Required | Description | Default |
|---|---|---|---|
| object | No | ||
| subject | Yes | ||
| relation | Yes |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
With no annotations, the description fully handles transparency. It discloses that deletion is permanent and provable, that facts are fully erased, and that a signed proof of deletion is returned. This is comprehensive for a destructive operation.
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, with key information front-loaded (permanence, GDPR). Every sentence adds value: purpose, usage, parameter hint, and output confirmation. 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 (3 params, no output schema, no annotations), the description covers the core purpose, usage context, and parameter behavior partially. It mentions a signed proof but does not describe the response format. Lacks explanation of required parameters, but overall is adequate for basic 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?
The input schema has 0% description coverage, so the description must compensate. It only explains the 'object' parameter's optionality (omit to delete all values), but does not define 'subject' or 'relation'. This leaves ambiguity about their meaning, which is insufficient for correct invocation.
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 that the tool permanently deletes a stored fact for GDPR compliance. The verb 'forget' is somewhat abstract, but the description clarifies its meaning. It distinguishes from sibling tools like learn_fact and update_fact by focusing on deletion.
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 explicitly says to use when the user asks to forget or remove information, providing clear context. It also mentions the optional behavior of omitting the object to delete all values, but does not specify when not to use the tool or alternatives.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Full life-line of a fact (all versions live+closed, with sources and valid-time). Audit/compliance.
| Name | Required | Description | Default |
|---|---|---|---|
| subject | Yes | ||
| relation | Yes |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
No annotations are provided, so the description must carry the full behavioral burden. It discloses returning all versions (live and closed) with sources and valid-time, but lacks details on read-only nature, authorization, performance, or potential side effects.
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 sentence that conveys the core purpose efficiently without unnecessary words or repetition.
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 tool with two required parameters and no output schema, the description omits important details such as return format, pagination, and parameter constraints. While the purpose is clear, the lack of parameter guidance and usage context makes it incomplete for an agent.
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 does not explain the parameters 'subject' and 'relation'. The tool expects two required string parameters with no additional meaning beyond their names, leaving the agent with no guidance on values or formats.
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 that the tool retrieves the full life-line of a fact, including all versions (live and closed), sources, and valid-time, explicitly for audit/compliance. This distinguishes it from sibling tools like 'prove_fact' or '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?
The description implies use for audit/compliance but does not explicitly state when to use or when alternatives are better. No exclusions or alternative tools are mentioned, leaving the agent to infer context.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Merkle root committing the entire current knowledge state (one hash).
| Name | Required | Description | Default |
|---|---|---|---|
No parameters | |||
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
The description implies a read-only operation (returning a hash) but does not explicitly state that it has no side effects or that it is safe. With no annotations, the description should be more explicit about the read-only nature and any potential costs or restrictions.
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, clear sentence with no filler. It is front-loaded and efficiently conveys the essential information.
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's simplicity (no parameters, no output schema), the description is mostly complete. It states the return value (one hash) and its purpose. However, it could elaborate on how this hash might be used or verified, but that is a minor gap.
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?
There are no parameters, so the schema fully covers requirements. The description does not need to add parameter details, but it could hint at the output format. A score of 4 is appropriate as the baseline for zero-parameter tools.
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 identifies the tool as providing a Merkle root hash of the entire knowledge state. It uses a specific noun (Merkle root) and implies a read operation. It is distinct from siblings like 'learn_fact' or '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?
No guidance is provided on when to use this tool versus alternatives. There is no mention of context or prerequisites, leaving the agent to infer usage from the name alone.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Store a fact (subject, relation, object) the agent must recall EXACTLY later, with its source. Call whenever the user states a fact, preference, decision, name, number, or rule worth remembering — it persists across sessions and is never silently distorted. Optional valid_from for valid-time.
| Name | Required | Description | Default |
|---|---|---|---|
| object | Yes | ||
| source | No | ||
| subject | Yes | ||
| relation | Yes | ||
| valid_from | No |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
No annotations are provided, so the description carries full burden. It adds value by stating 'never silently distorted' and implicit persistence. However, it fails to disclose side effects on duplicate facts, required permissions, or error 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, no filler. First sentence defines purpose, second sentence provides usage context and mentions optional parameter. Every word 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?
Covers basic usage and parameters. However, it does not explain return values (no output schema required) nor guide the agent to sibling tools like recall or update_fact. For a knowledge management tool among many siblings, more integration context is needed.
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 description must compensate. It identifies all parameters (subject, relation, object, source, valid_from) and explains object as 'fact', source as 'with its source', and valid_from as 'optional for valid-time'. This adds meaning beyond the bare 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 'Store a fact (subject, relation, object) the agent must recall EXACTLY later, with its source.' It defines the verb (store) and resource (fact). However, it does not differentiate from sibling tools like update_fact, leaving ambiguity about whether it overwrites or appends.
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 says 'Call whenever the user states a fact, preference, decision, name, number, or rule worth remembering.' This gives clear when-to-use guidance. But it lacks when-not-to-use or mention of alternative tools like update_fact for modifications.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Multi-hop chain: start entity + list of relations, follows subject->object each step (exact only, 0% hallucination).
| Name | Required | Description | Default |
|---|---|---|---|
| as_of | No | ||
| start | Yes | ||
| relations | Yes |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
With no annotations, the description must carry full burden for behavioral traits. It mentions 'exact only, 0% hallucination' which addresses accuracy, but omits details on side effects, authentication needs, rate limits, or whether it is read-only.
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 sentence, concise and front-loaded with the main action. However, it could be better structured with bullet points or clear separation of purpose, parameters, and behavior.
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 of a multi-hop chain tool and the absence of an output schema, the description is insufficient. It lacks information on return format, error handling, path length limits, or edge cases.
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 explains 'start' as start entity and 'relations' as list of relations, but fails to describe the 'as_of' parameter, format details, or allowed values. Two of three parameters are partially explained.
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 performs a multi-hop chain: starting from an entity and following a list of relations step by step. This distinguishes it from sibling tools like 'all_paths' (likely all paths) and 'prove_fact' (single fact).
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 any guidance on when to use this tool versus alternatives. It lacks context on prerequisites, preferred scenarios, or comparisons to other sibling tools.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Merkle inclusion proof that a fact is in the knowledge state (without revealing other facts).
| Name | Required | Description | Default |
|---|---|---|---|
| subject | Yes | ||
| relation | Yes |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
With no annotations, the description must carry behavioral info. It mentions 'Merkle inclusion proof' and privacy, but does not disclose error behavior, performance implications, or edge cases (e.g., what happens if the fact is absent).
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 that is front-loaded and contains no superfluous words. Every part adds value.
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 2-param tool with no output schema or annotations, the description is adequate but limited. It doesn't explain the output format (the proof) or how to use it with siblings like 'verify_proof'.
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 description must compensate. However, it only names 'subject' and 'relation' indirectly without explaining their meaning or adding any constraints 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 specifies that the tool generates a Merkle inclusion proof for a fact in the knowledge state, with an added privacy guarantee ('without revealing other facts'). This distinguishes it from sibling tools like 'learn_fact' and '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?
The description implies a use case (proving existence without revealing other facts) but does not explicitly state when to use this tool over alternatives, or any prerequisites or exclusions.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Look up a stored fact and return the answer WITH its cited source — or an honest 'unknown'. ALWAYS call this before answering a factual or memory question instead of guessing: it returns nothing rather than hallucinating, and includes a signed, verifiable receipt. Optional as_of for valid-time.
| Name | Required | Description | Default |
|---|---|---|---|
| as_of | No | ||
| subject | Yes | ||
| relation | Yes |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
No annotations are provided, so the description fully carries the burden. It discloses that the tool returns answers with cited sources or 'unknown', includes a signed receipt, and avoids hallucination. It also mentions the optional 'as_of' parameter for valid-time queries, covering key behavioral traits.
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 (three sentences), with the first sentence immediately stating the core function, the second providing usage guidance, and the third adding a parameter note. No unnecessary 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 a simple lookup tool with no output schema and 3 parameters, the description covers the return format (answer with source or 'unknown', receipt), usage context, and valid-time capability. It is sufficiently complete, though it could briefly clarify the role of subject and relation.
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%, requiring the description to explain parameters. It only explains 'as_of' (optional for valid-time), but does not describe the required 'subject' and 'relation' parameters, leaving their meanings implicit. Users must infer that they specify the fact to be recalled.
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: to look up stored facts and return answers with cited sources or 'unknown'. It distinguishes itself from sibling tools by emphasizing honest recall over guessing and providing verifiable receipts.
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 explicitly advises to always call this tool before answering factual or memory questions instead of guessing, and notes that it returns nothing rather than hallucinating. While it does not enumerate alternatives, the context of sibling tools implies when not to use (e.g., for learning or forgetting).
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Counts + current knowledge root.
| Name | Required | Description | Default |
|---|---|---|---|
No parameters | |||
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
No annotations provided, so description carries full burden. It only states return values but does not disclose read-only nature, side effects, or required permissions. Essential behavioral traits are missing.
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 (one sentence) but sacrifices clarity. It could be longer to explain what 'counts' means. It is not front-loaded with key information.
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 parameters, the description must explain return format. It fails to specify what counts are, leaving the agent guessing. Incomplete for effective 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?
No parameters exist, so schema coverage is 100%. The description adds no parameter info, but baseline is high due to absence of parameters. No value lost.
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 'Counts + current knowledge root' names two pieces of return data but is vague about what counts (facts? nodes?). It partially distinguishes from sibling 'knowledge_root' by adding counts, but lacks specificity about the counts.
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 guidance on when to use this tool vs siblings like 'knowledge_root' or 'history'. The description does not mention prerequisites, exclusions, or alternative tools.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Update a fact's value without retraining: closes live versions (valid_to=t) and opens a new one. History is preserved.
| Name | Required | Description | Default |
|---|---|---|---|
| t | No | ||
| source | No | ||
| subject | Yes | ||
| relation | Yes | ||
| new_object | Yes |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
With no annotations provided, the description discloses the key behavioral trait: closing old versions and opening a new one while preserving history. It lacks details on error conditions or required permissions, which would raise it to 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?
Every word serves a purpose. Two sentences efficiently convey the action, mechanism, and side effect. Front-loaded with core purpose.
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 core behavior but lacks parameter guidance and return value info (no output schema). For a 5-parameter mutation tool, more context is needed for safe 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%, and the description provides no explanation of any of the 5 parameters (t, source, subject, relation, new_object). The agent must infer parameter roles solely from parameter 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?
The description clearly states the verb (update), the resource (fact), and the mechanism (closes live version, opens new one) while preserving history. It distinguishes from sibling tools like 'learn_fact' (add new) and 'forget' (delete).
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?
Implied usage context ('without retraining') differentiates from retraining workflows. However, it does not explicitly state when not to use this tool or suggest alternatives among siblings for related tasks.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Verify a Merkle inclusion proof (leaf, proof, root).
| Name | Required | Description | Default |
|---|---|---|---|
| leaf | Yes | ||
| root | Yes | ||
| proof | Yes |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
No annotations provided, and the description only states the action without detailing side effects, return values, or behavior on success/failure.
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 very short and front-loaded with the key information, containing no unnecessary 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?
Lacks details on return values, error conditions, or how the proof is validated, which are critical for a verification tool with no output schema.
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%. The description mentions parameter names but does not explain their meaning, format, or constraints.
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 identifies the tool as verifying a Merkle inclusion proof and lists the required parameters. However, it does not differentiate from sibling tools like 'prove_fact' or 'verify_receipt'.
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 guidance on when to use this tool versus alternatives, nor any context about prerequisites or exclusions.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Verify a signed receipt returned by recall/forget (tamper-evident).
| Name | Required | Description | Default |
|---|---|---|---|
| receipt | Yes |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
With no annotations, the description carries the full burden. It mentions 'verify' and 'tamper-evident', suggesting authenticity checks, but does not disclose verification outcomes (e.g., success/failure format), error conditions, or any side effects.
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 redundant words. Every word adds value: verb, resource, source, and key property.
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 of the tool (one parameter, no output schema), the description is mostly complete. It could specify return values (e.g., boolean or decoded receipt) but the core purpose and parameter context are sufficient.
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 single parameter 'receipt' is an object with no schema description. The description adds meaningful context: it is a 'signed receipt returned by recall/forget'. This compensates for the 0% schema coverage by linking the parameter to its origin.
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 verb 'verify' and the resource 'signed receipt', and specifies the source ('returned by recall/forget') with a notable property ('tamper-evident'). This distinguishes it from siblings like recall and forget, which produce receipts.
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 use after recall/forget to check integrity but does not explicitly state when to use or not use this tool, nor mention alternatives. It lacks guidance on prerequisites or fallback scenarios.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Recent tool additions, removals, and schema changes observed during successful MCP inspections. Dates show when Glama detected each change.
v0.1.2all_pathscontradictionsforgethistoryknowledge_rootlearn_factmultihopprove_factrecallstatsupdate_factverify_proofverify_receiptEvery tool has a clearly distinct purpose: storing (learn_fact), recalling (recall), updating (update_fact), deleting (forget), auditing (history, contradictions), proving (prove_fact, verify_proof, verify_receipt), and traversing (multihop, all_paths). No two tools overlap in functionality.
All tool names use consistent snake_case with a verb_noun pattern (e.g., learn_fact, update_fact, prove_fact) or noun_verb (knowledge_root, verify_receipt). The naming is predictable and intuitive.
With 13 tools, the server covers the essential operations for a verifiable memory system without being excessive. Each tool justifies its existence by enabling a distinct workflow (CRUD, proofs, audits, traversal).
The tool surface is comprehensive: it includes create (learn_fact), read (recall, history, contradictions), update (update_fact), delete (forget), proof generation and verification (prove_fact, verify_proof, verify_receipt), multi-step traversal (multihop, all_paths), and state commitment (knowledge_root, stats). No obvious gaps.
Unclaimed servers have limited discoverability.
If you are the server author, to access and configure the admin panel.
Persistent, inspectable memory for AI agents with lineage, correction, and a hosted MCP endpoint.
Agent memory that refuses to guess: evidence-gated recall, exact-source reads, verifiable deletion.
Persistent memory for AI agents. Semantic search, memory graph, W3C DID identity.
Persistent memory for AI agents. EU-hosted, privacy-first, hybrid recall, contradiction detection.
We provide all the information about MCP servers via our MCP API.
curl -X GET 'https://glama.ai/api/mcp/v1/servers/Mars-proj/verifiable-memory'
If you have feedback or need assistance with the MCP directory API, please join our Discord server
