Click on "Install Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type @ followed by the MCP server name and your instructions, e.g., "@safehold provide my passport number for flight booking"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
Your identity documents, on your machine, shared only when you say so.
Safehold is a local MCP server that stores passports, visas, driving licences, and other identity documents. When an AI agent needs your passport number to book a flight, it asks Safehold. You see what's being requested and why, then approve or deny. Nothing leaves your device.
Related MCP server: aip-identity
Your identity data is scattered across dozens of services you don't control: banks, airlines, government portals, SaaS products. Each one holds a copy and decides for itself how to store it, whether to sell it, and how long it'll stick around. AI agents make this worse. Anything an agent can access, every API it touches can reach too.
Safehold is the alternative. The agent gets a field value. You get a consent prompt instead of a privacy policy. Your files stay where they are.
Open source, so none of this requires trust.
A few things that get confused with safehold but solve a different problem.
You can stuff a passport into 1Password. The op CLI can read specific fields. There's even an official 1Password MCP server now. So yes, parts of this overlap.
What's different is the trust model. Once your 1Password vault is unlocked, anything with that session can read anything in it. No per-field consent, no "why are you asking for this?" prompt. The vault is built for humans typing passwords, not for agents pulling structured fields with a stated reason.
Safehold gates each field individually, every time, with a purpose attached.
Works fine if your agent is a browser filling a form. Doesn't help when it isn't. Agents working over APIs, CLI tools, or other MCP servers never see the autofill layer, so you end up copy-pasting from a passwords app into a chat anyway. Safehold gives a non-browser agent a way to request a passport number without that detour.
Locked to the platform. Apple Wallet won't hand your boarding pass to a third-party agent. mDL pilots and the EU eIDAS wallet are issuer-controlled and narrow in scope. None of them speak MCP, and none will let an agent on your laptop request a field from your driving licence.
People do this. They drop their passport number into a system prompt, or type it into a message and trust the agent to handle it. The number then lives in chat history, gets shipped to the model provider, and sticks around in their logs. You also have no record of which agent used it for what.
Safehold keeps the document on disk, encrypted, and only releases the one field the agent asked for. Nothing extra ends up in the conversation.
Tempting if you're a developer. ~/identity.json plus the filesystem MCP is three lines of config and done. It also gives every other agent on the box the same access, with no encryption at rest, no logging, and no consent step. The first time that file shows up in a backup or a screen share, you have a problem.
KYC vendors. You send them your documents, they store them, and partner services query them on your behalf. The model assumes you don't hold the keys. Safehold takes the other direction: documents stay on your machine and you approve every release.
Built for service secrets, not personal identity. No human-facing consent flow, no schema for documents, and operationally heavy for someone who just wants their passport readable by an agent on their laptop.
Safehold runs as a stdio MCP server. Documents go into ~/.safehold/vault/ as individual encrypted files. Every retrieval is logged to ~/.safehold/access-log.enc.
Encryption is XChaCha20-Poly1305 via @noble/ciphers, audited by Cure53, no external dependencies. On first run a 32-byte key is written to ~/.safehold/master.key (mode 0600).
Other MCP clients can request specific fields from your documents. You control what gets shared, every time.
Tool | Description |
| Add a document to the vault |
| List stored documents (metadata only) |
| Update fields on an existing document |
| Permanently delete a document |
| Retrieve specific fields from a passport |
| Retrieve specific fields from a national ID |
| Retrieve specific fields from a driving licence |
| Retrieve specific fields from a visa |
| Retrieve a stored photo (passport style, selfie, or signature) |
| Retrieve fields from any document type |
| View what was accessed, by which client, and when |
Every retrieval tool accepts a purpose parameter. That purpose is shown to you at consent time so you always know why a field is being requested.
Safehold also ships ready-made workflow prompts. These show up as slash commands in MCP clients (Claude Desktop, Cursor, Claude Code, Zed) so common tasks don't have to be re-prompted from scratch.
Prompt | Description |
| Guided intake to add or update a passport, national ID, driving licence, or visa |
| Check passport and visa validity for an upcoming trip |
| Share the minimum a service actually needs, with an explicit consent purpose |
| Surface anything expiring soon (metadata only, no consent prompts) |
Prompts don't access your data themselves. They tell the calling agent which Safehold tools to call, in what order, and with what purpose. Every retrieval still goes through the normal consent flow.
safehold-skills is a companion package of Claude Code skills that turn Safehold from a vault into finished workflows: airline check-in, ESTA applications, KYC forms, and so on. Each skill pairs Safehold with a browser MCP and asks for fields one at a time with a precise purpose string.
npx safehold-skills installRequires Node.js >=20. No build step.
Most clients use the same JSON config — just the file path differs:
{
"mcpServers": {
"safehold": {
"command": "npx",
"args": ["-y", "safehold@latest"]
}
}
}Client | Config file |
Claude Desktop (macOS) |
|
Claude Desktop (Windows) |
|
Cursor (project) |
|
Cursor (global) |
|
Windsurf |
|
Amazon Q Developer (global) |
|
Amazon Q Developer (project) |
|
Clients with slightly different config formats:
Claude Code (CLI)
claude mcp add safehold npx -- -y safehold@latestVS Code (GitHub Copilot) — uses "servers" key, enable Agent mode in Copilot Chat. Edit .vscode/mcp.json (project) or ~/Library/Application Support/Code/User/mcp.json (global, macOS):
{
"servers": {
"safehold": {
"command": "npx",
"args": ["-y", "safehold@latest"]
}
}
}Zed — uses "context_servers" key in ~/.config/zed/settings.json:
{
"context_servers": {
"safehold": {
"command": "npx",
"args": ["-y", "safehold@latest"]
}
}
}OpenAI Codex CLI — TOML format in ~/.codex/config.toml:
[mcp_servers.safehold]
command = "npx"
args = ["-y", "safehold@latest"]Continue.dev — create .continue/mcpServers/safehold.json (enable Agent mode):
{
"command": "npx",
"args": ["-y", "safehold@latest"]
}Hermes (Nous Research) — YAML format in ~/.hermes/config.yaml:
mcp_servers:
safehold:
command: npx
args: ["-y", "safehold@latest"]OpenClaw — uses "servers" key in ~/.openclaw/mcp.json:
{
"servers": {
"safehold": {
"command": "npx",
"args": ["-y", "safehold@latest"]
}
}
}get_travel_authorization - ESTA, eTA, UK ETA, and similar
get_vaccination_certificate - international vaccination records
get_residency_permit - BRP, Green Card, and equivalents
get_birth_certificate
get_tax_id - NI number, SSN, TFN, and equivalents
get_proof_of_address - utility bills, bank letters
get_travel_insurance
npm install
npm test # run tests
npm run typecheck # type check
npm run build # compile to dist/
npm run dev # run directly with tsx (no build step)npm version patch # bug fix (0.1.0 → 0.1.1)
npm version minor # feature (0.1.0 → 0.2.0)
npm version major # breaking (0.1.0 → 1.0.0)Checks that you're on a clean, up-to-date main branch, bumps the version, commits, tags, and pushes. CI publishes to npm.
Add a new document to the vault. The document is stored locally and encrypted. Known document types (passport, national_id, driving_license, visa, photo_*) are strictly validated.
| Name | Required | Description | Default |
|---|---|---|---|
| label | Yes | A human-readable label for this document, e.g. "UK Passport 2024" | |
| fields | Yes | The document field values to store | |
| document_type | Yes | The type of document (e.g. "passport", "visa") |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
With no annotations provided, the description carries the burden of disclosing behavior. It discloses that documents are stored locally, encrypted, and strictly validated for known types. This is valuable context beyond the tool's basic purpose, though it omits return behavior and authentication 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?
The description is a concise three-sentence structure that front-loads the primary purpose. Every sentence adds useful information: what it does, how it stores, and validation rules. No fluff or redundancy.
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 the core operation, storage, encryption, and validation, but it lacks details on expected return values, failure behavior, and idempotency. For a create operation with no output schema, these details would be helpful for an agent to confidently invoke 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 covers 100% of parameters, so the baseline is 3. The description adds meaning beyond the schema by specifying that known document types (passport, national_id, etc.) are strictly validated, which clarifies the semantics of the document_type parameter.
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 explicitly states 'Add a new document to the vault', which is a specific verb+resource action. It clearly distinguishes from sibling tools like update_document and delete_document by emphasizing 'new'.
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 clearly indicates this is for adding new documents, but it does not explicitly state when not to use it (e.g., 'use update_document for existing documents'). The usage context is implied by 'new' but lacks explicit exclusions or alternatives.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Permanently delete a document from the vault.
| Name | Required | Description | Default |
|---|---|---|---|
| id | Yes | The ID of the document to delete |
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 explicitly states the deletion is permanent, which adds a mild warning, but it does not disclose potential side effects, authorization needs, or what happens to associated data. The 'permanently' is the only behavioral trait mentioned.
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, concise sentence that is front-loaded with the essential action and object. Every word contributes meaning, with no unnecessary 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 destructrive tool, the description is minimally adequate: it states the action and the target. However, it lacks usage guidance and any context about consequences or expected outcomes, which would be expected for a mutation tool even without an 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?
The schema already provides full coverage of the single parameter ('id') with a clear description. The tool description adds no additional meaning beyond what the schema provides, so the baseline score of 3 applies.
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 ('delete') and resource ('document') plus scope ('from the vault'). This unambiguously distinguishes it from sibling tools such as get_document, add_document, and update_document.
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 are no mentions of prerequisites, conditions, or exclusions. The description solely states what it does without contextual usage cues.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
View the history of document field accesses — what was shared, with which client, and when.
| Name | Required | Description | Default |
|---|---|---|---|
| limit | No | Maximum number of log entries to return (default 50) | |
| document_id | No | Filter entries to a specific document ID |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
No annotations are provided, so the description carries full responsibility for disclosing behavioral traits. It mentions viewing access history but does not explicitly state that it is read-only, what permissions are required, whether only successful accesses are logged, or how results are ordered/paginated. This leaves significant ambiguity beyond the basic purpose.
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, well-structured sentence that immediately conveys the tool's function without unnecessary words. It front-loads the key action ('View the history') and then adds clarifying details.
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 tool has no output schema, and the description only gives a high-level summary of log entry content. It does not mention details like default limit (though in schema), pagination, ordering, or whether the log includes failed attempts. Given the tool's moderate complexity, the description is adequate but leaves gaps in what a complete response might look like.
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 fully describes both parameters (limit and document_id) with clear descriptions and defaults. The tool description adds no additional meaning or context about how these parameters affect results. Since schema coverage is 100%, the baseline score of 3 applies.
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 view the history of document field accesses. It specifies what the log contains (what was shared, with which client, and when), and this differentiates it from sibling tools like get_document or list_documents, which focus on retrieving documents themselves.
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 usage for auditing or reviewing access history, but does not explicitly state when to choose this tool over alternatives. While no sibling tool serves the same function, there is no explicit guidance on scenarios or prerequisites, so the usage context is only implied.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Retrieve fields from any document type not covered by a dedicated tool. The user will be prompted for consent before any data is returned.
| Name | Required | Description | Default |
|---|---|---|---|
| fields | Yes | The field names to retrieve from the document | |
| purpose | Yes | Why the calling agent needs this data — shown to the user for consent | |
| document_id | No | Specific document id (required when more than one of this type is stored) | |
| document_type | Yes | The document type to retrieve (e.g. "insurance_card") |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
The description discloses that the user will be prompted for consent before data is returned—a critical behavioral detail not present in annotations or schema. It adds transparency about the interactive nature of the operation, though it does not elaborate on error conditions or return format.
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 focused sentences, with the primary purpose stated first and the consent behavior following. No redundant details.
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 generic retrieval tool with no output schema or annotations, the description covers the essential context: what it does, when to use it (non-dedicated types), and a key behavioral note (consent prompt). It lacks explicit return format details, but the 'fields' concept and schema provide adequate information for a straightforward read operation.
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?
All four parameters are fully described in the schema (100% coverage), and the description does not add meaningful parameter-level detail beyond what is already available. It slightly reinforces document_type usage but does not exceed the baseline.
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 action (retrieve fields), the resource (any document type), and explicitly differentiates from dedicated getters by stating 'not covered by a dedicated tool,' distinguishing it from siblings like get_passport or get_visa.
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 'not covered by a dedicated tool' provides an explicit when-to-use condition and implies that dedicated tools should be used for their respective document types, thus giving clear guidance relative to siblings.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Retrieve specific fields from a stored driving licence. The user will be prompted for consent before any data is returned.
| Name | Required | Description | Default |
|---|---|---|---|
| fields | Yes | The driving licence fields to retrieve | |
| purpose | Yes | Why the calling agent needs this data — shown to the user for consent | |
| document_id | No | Specific driving licence id (required when more than one is stored) |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
With no annotations provided, the description carries the full burden and does disclose a key behavioral trait: the user will be prompted for consent before data is returned. This is meaningful context, though it does not cover other behaviors like error handling or the document_id requirement.
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 every sentence carries essential information (what it does and the consent behavior). No redundancy or 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?
The tool is moderately complex with three parameters and no output schema. The description covers the consent behavior but omits practical details like the document_id requirement for multiple stored licences (present only in the schema) and what happens if consent is denied. The schema helps, but the description alone is not 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?
Schema coverage is 100%, so the baseline is 3. The description's mention of 'specific fields' aligns with the fields parameter but adds no new semantics beyond the schema, which already describes the fields and purpose parameters clearly.
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 specific fields from a stored driving licence, using a specific verb and resource. It distinguishes itself from sibling tools like get_passport and get_national_id by naming the driving licence specifically.
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 usage for retrieving driving licence fields but provides no explicit guidance on when to use this tool versus alternatives like get_document or when not to use it. The consent note gives context but no comparative guidance.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Retrieve specific fields from a stored national identity card. The user will be prompted for consent before any data is returned.
| Name | Required | Description | Default |
|---|---|---|---|
| fields | Yes | The national ID fields to retrieve | |
| purpose | Yes | Why the calling agent needs this data — shown to the user for consent | |
| document_id | No | Specific national ID id (required when more than one is stored) |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
There are no annotations, so the description carries the disclosure burden. It does disclose the important behavioral trait that the user is prompted for consent before data is returned, which is valuable. However, it omits other behavioral details such as behavior on consent denial, missing document, or when multiple national IDs exist and document_id is required.
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 concise sentences, front-loaded with the action and resource, and every clause adds value. The consent sentence is essential behavioral context and is not redundant.
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 retrieval tool with no output schema, the description is mostly sufficient: it states what is retrieved, that fields are specific, and that consent is required. It could be slightly more complete by describing the return shape or failure behavior, but the essential invocation context is present.
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 the schema already explains the fields, purpose, and document_id parameters. The description reinforces 'specific fields' and consent but adds no new meaning beyond the schema, such as value formatting, enumeration semantics, or edge-case behavior for optional document_id.
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 ('Retrieve') and resource ('stored national identity card'), and adds the scope 'specific fields', distinguishing it from generic get_document and identity-specific siblings like get_passport and get_driving_license. The consent note further clarifies the core action.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
No explicit when-to-use or when-not-to-use guidance is provided, nor are sibling alternatives mentioned. However, the intended context is implied clearly: use this tool to pull selected fields from a national ID card. It lacks exclusions or comparison with get_document or other identity getters.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Retrieve specific fields from a stored passport. The user will be prompted for consent before any data is returned.
| Name | Required | Description | Default |
|---|---|---|---|
| fields | Yes | The passport fields to retrieve | |
| purpose | Yes | Why the calling agent needs this data — shown to the user for consent | |
| document_id | No | Specific passport id (required when more than one passport is stored) |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
With no annotations provided, the description carries the full burden. It discloses a key behavior: the user will be prompted for consent before data is returned. However, it does not mention what happens if consent is denied, whether the operation is read-only (implied), or any other side effects or limitations. The consent disclosure is valuable but not exhaustive.
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 main purpose, and every phrase earns its place. The consent prompt sentence adds crucial behavioral information without 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?
For a simple read tool, the description covers the core purpose and consent behavior. It does not explicitly mention the document_id requirement (though the schema handles that) or differentiate itself from the generic get_document sibling. Given no output schema, the description could be slightly richer, but it is still adequate for the tool's complexity.
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 100% (all three parameters have descriptions), so the baseline is 3. The description adds minimal semantic value by mentioning 'specific fields' (relating to the fields array) and consent (relating to the purpose string), but does not elaborate 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 ('Retrieve') and resource ('stored passport') and clearly scopes to 'specific fields'. This distinguishes it from sibling tools like get_national_id, get_driving_license, and get_visa, which target other document types, and from the generic get_document.
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 makes clear this is for retrieving specific passport fields and mentions the consent prompt, giving context on when to use it. However, it lacks explicit exclusions or alternative guidance (e.g., when to use get_document instead), but the context is sufficiently clear for a passport-specific tool.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Retrieve a stored photo. Returns base64-encoded image data. The user will be prompted for consent before any data is returned.
| Name | Required | Description | Default |
|---|---|---|---|
| type | Yes | The type of photo to retrieve | |
| purpose | Yes | Why the calling agent needs this photo — shown to the user for consent | |
| document_id | No | Specific photo id (required when more than one of this type is stored) |
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 informs the agent of the base64 return format and the user consent prompt, which are important behavioral traits. It does not mention error cases or side effects, but for a retrieval 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 three concise sentences, each carrying essential information: the action, the return format, and the consent requirement. There is no fluff or redundancy, making it highly effective.
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 retrieval tool with full schema coverage, the description adequately covers the key behavioral notes (base64 format and consent). It does not explicitly mention conditional requirements like when document_id is needed, but the schema covers that, making the description complete enough.
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 100%, so each parameter is already documented in the schema. The description adds no additional meaning beyond what the schema provides, such as specifics about the type or purpose fields, so the baseline 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 'Retrieve a stored photo' with a specific verb and resource, making the tool's function unambiguous. However, it does not explicitly distinguish itself from sibling tools like get_document or get_passport, so it lacks sibling differentiation.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
No guidance is provided on when to use this tool versus alternatives such as get_passport or get_document. The consent note implies a usage context but does not offer explicit exclusions or comparison with other retrieval tools.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Retrieve specific fields from a stored visa. The user will be prompted for consent before any data is returned.
| Name | Required | Description | Default |
|---|---|---|---|
| fields | Yes | The visa fields to retrieve | |
| purpose | Yes | Why the calling agent needs this data — shown to the user for consent | |
| document_id | No | Specific visa id (required when more than one visa is stored) |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
With no annotations provided, the description carries the burden of behavioral disclosure. It explicitly discloses a key privacy-relevant behavior: 'The user will be prompted for consent before any data is returned.' This goes beyond the schema and is critical for an agent to set expectations. It does not cover other behaviors (e.g., what happens if consent is denied), but the consent note adds substantial value.
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 exceptionally concise: two short sentences that front-load the core purpose and then add the essential consent disclosure. Every word earns its place, with no redundancy or irrelevant details.
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 retrieval tool with detailed schema and a consent note, the description is largely complete. It covers the main action and a critical behavioral aspect, but it omits mention of what happens if the user denies consent or if the document_id is missing when multiple visas exist (though the latter is in the schema). Given the tool's low complexity and rich schema, these gaps are acceptable but not negligible.
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 100%, so the schema fully documents each parameter. The description adds only a slight nuance by referencing 'specific fields' and linking 'purpose' to consent, but it does not elaborate on parameter syntax or values beyond what the schema already provides. This meets the baseline for schema-heavy descriptions.
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 ('Retrieve specific fields') applied to a specific resource ('a stored visa'), which distinguishes it from sibling tools like get_passport and get_national_id by naming the visa type and the selective nature of the retrieval. This makes the tool's function 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 when you need a subset of visa data, but it does not explicitly state when to choose this over get_document or list_documents, nor does it mention any exclusions. The 'specific fields' phrasing hints at selective retrieval, but no alternatives or when-not-to-use guidance is given.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
List all stored documents. Returns metadata only (type, label, expiry) — no sensitive field values.
| Name | Required | Description | Default |
|---|---|---|---|
No parameters | |||
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
With no annotations, the description carries full burden. It discloses the return scope (all documents), the exact fields returned (type, label, expiry), and explicitly states sensitive field values are not included. This provides meaningful behavioral context beyond the tool name, though it omits details like authentication or pagination.
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, front-loaded sentence states the action, then adds a qualifying clause about return content. Every word earns its place; no redundancy or fluff.
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 zero-parameter list tool with no output schema, the description fully covers what it does and what it returns. It specifies the metadata fields and explicitly notes absence of sensitive values, which is sufficient context for an agent to decide when to use 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?
There are 0 parameters, so a baseline of 4 is appropriate. The description adds no parameter information, but none is needed; it clarifies the tool takes no filters, which aligns with the empty 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 specific verb+resource: 'List all stored documents.' It clearly distinguishes from siblings like get_document by specifying it returns metadata only, not sensitive field values.
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?
Clear context: this tool lists all documents, implying use for overviews. It implicitly differentiates from get_document by noting sensitive values are excluded, but it does not explicitly name alternatives or state 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.
Update fields on an existing document. Only the provided keys are changed. Known document types are strictly validated.
| Name | Required | Description | Default |
|---|---|---|---|
| id | Yes | The ID of the document to update | |
| fields | Yes | The fields to update — only provided keys are changed |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
The description discloses two key behaviors: partial updates and strict validation of known document types. However, with no annotations, it omits important context such as side effects, permissions, error handling, or reversibility. It adds some value but falls short of full transparency.
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 concise sentences, front-loaded with the purpose and followed by a key behavioral detail. Every word earns its place and there is no redundant 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?
With no output schema and no annotations, the description should clarify return values, error scenarios, and relationship to specific document-type tools. It partially does so but remains incomplete for a mutation tool, especially given the sibling tools that suggest varied document types.
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 100% for both parameters, so the schema already documents their meaning. The description repeats the partial-update behavior that is already present in the fields parameter description, adding no new parameter-specific 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 specifies a clear verb-resource pair: 'Update fields on an existing document.' It also adds detail with 'Only the provided keys are changed,' which clarifies partial-update semantics. This distinguishes it from sibling tools like add_document and delete_document.
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 usage for modifying existing documents but does not explicitly state when to use it versus alternatives. No exclusions or alternative tool references are given, so guidance remains implicit rather than explicit.
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.
v1.2.0add_documentdelete_documentget_access_logget_documentget_driving_licenseget_national_idget_passportget_photoget_visalist_documentsupdate_documentMost tools have clearly distinct purposes (get_passport vs get_national_id, etc.), but the generic get_document overlaps with the dedicated getters, potentially causing selection confusion when a specific document type is already known.
All tool names follow a consistent verb_noun pattern in snake_case (list_, get_, add_, update_, delete_), with no mixed conventions or vague verbs.
With 11 tools, the set is well-scoped for a document vault: listing, retrieving by type, adding, updating, deleting, and access logging. Each tool earns its place, and the count feels neither thin nor bloated.
The tool surface covers the full document lifecycle (create, read, update, delete, list) and adds an access log for auditing. Specific getters for common document types plus a generic fallback ensure no obvious dead ends.
Unclaimed servers have limited discoverability.
If you are the server author, to access and configure the admin panel.
MCP server for verifying EUDI/Talao wallet data via OIDC4VP (pull) for AI agents.
Governed personal world model and memory for your AI agent. Pair once, connect over MCP.
AI governance MCP server for EU AI Act compliance and jurisdiction verification
Persistent memory for AI agents with OAuth-backed hosted MCP access.
We provide all the information about MCP servers via our MCP API.
curl -X GET 'https://glama.ai/api/mcp/v1/servers/Spacetime-Technology/safehold'
If you have feedback or need assistance with the MCP directory API, please join our Discord server
