Zero-retention GDPR PII redaction gate (IBAN, Tax-ID, HRB, Names) with signed Ed25519 VC.
Scored across 3 tools
Each tool targets a clearly different compliance operation: PII redaction, sanctions/KYB screening, and e-invoice XML validation. There is no plausible way an agent would confuse one for another, since the names and descriptions name distinct artifacts and standards.
All three follow a verb-first snake_case pattern (redact_, screen_, validate_) with domain-specific objects. Minor deviations such as the conjunction in 'screen_sanctions_and_kyb' and the redundant 'gdpr_pii' stacking keep it from perfect predictability, but the convention is readable and stable.
Three tools is thin for a suite billed as a GDPR/AI Act compliance gate; the surface is essentially one call per regulation artifact. Each tool is meaningful, but the count is at the borderline where the set feels more like a sampler than a complete gate.
All three tools issue signed Ed25519 VCs, yet there is no tool to verify or retrieve those credentials, no audit/log listing, and no AI Act-specific risk classification or pseudonymization distinct from redaction. These gaps will force agents to dead-end when they need to validate or look up previously issued attestations.
Zero-retention GDPR PII redaction gate (IBAN, Tax-ID, HRB, Names) with signed Ed25519 VC.
| Name | Required | Description | Default |
|---|---|---|---|
| raw_text | Yes | Unstructured text containing potential DACH/EU PII. | |
| max_accepted_fee_eur | No | ||
| ap2_intent_mandate_id | No |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
With no annotations, the description carries the full behavioral burden. It usefully discloses zero-retention processing and a signed Ed25519 VC, but it does not explain whether raw text is transformed or blocked, what the signed VC contains, authentication needs, or how the result is returned.
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 phrase with no wasted sentences, so it is concise. It is somewhat dense with acronyms (IBAN, HRB, VC), which slightly reduces clarity, but the core purpose comes first.
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 three parameters, no annotations, 33% schema coverage, and no output schema, the definition is too thin. It omits the roles of two parameters and does not clarify the return behavior of the signed VC, leaving meaningful gaps for an agent to call it correctly.
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 only 33%, and the description does not compensate. raw_text is described in the schema and its PII categories are implied in the description, but max_accepted_fee_eur and ap2_intent_mandate_id receive no meaning in either place.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description states a specific verb and resource: GDPR PII redaction on text, with named PII categories and a signed Ed25519 VC. It is clear enough to distinguish from the sanctions/KYB and e-invoice siblings by domain, but it does not explicitly differentiate itself from them.
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?
There is no explicit guidance on when to use this tool versus alternatives, nor when not to use it. The word 'gate' weakly implies preprocessing, but the description leaves the agent to infer the workflow context.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Deterministic EU FSF, UN & OpenSanctions KYB screening issuing signed Ed25519 Non-Hit VCs.
| Name | Required | Description | Default |
|---|---|---|---|
| entity_name | Yes | Legal company or beneficiary name to screen. | |
| country_code | No | ISO-3166-1 alpha-2 country code. | DE |
| vat_or_lei_id | No | Optional EU VAT ID or LEI identifier. | |
| max_accepted_fee_eur | No | ||
| ap2_intent_mandate_id | No |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
No annotations exist, so the description carries the full burden. 'Deterministic' and 'issuing signed Ed25519 Non-Hit VCs' usefully signal that the result is a cryptographically signed credential, but it never states what happens on a hit (presumably no VC is issued), whether calls are billed (the fee parameter implies cost), or whether screening state is stored. These are material behaviors for a compliance tool with no structured disclosure.
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 with no wasted filler, which is good, but the density of unexplained acronyms (FSF, KYB, VC, AP2) makes it cryptic rather than efficient. Brevity here trades against comprehension.
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 annotations and no output schema, the description must explain result shape, hit-vs-non-hit behavior, and cost. It hints at the credential output but omits outcome semantics and billing, leaving an agent unable to predict what a call will do or return in the negative case.
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 60%, leaving max_accepted_fee_eur and ap2_intent_mandate_id with no description anywhere. The description adds zero parameter meaning and is silent on the fee ceiling, which is the economically significant control, and on what an AP2 intent mandate is required for.
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 names a specific verb and resource ('KYB screening') and enumerates the exact sanctions lists screened (EU FSF, UN, OpenSanctions), plus the artifact produced (signed Ed25519 Non-Hit VCs). It is clearly distinct from the unrelated siblings (GDPR redaction, e-invoice validation). Domain jargon ('VCs', 'FSF', 'KYB') keeps it from being fully self-evident to a general agent.
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?
There is no statement of when to use this tool, when not to, or what the prerequisites are (e.g., a mandate ID requirement). Siblings are in completely different domains so no differentiation is needed, but the agent is given no situational triggers.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Official KoSIT EN 16931 & XRechnung XML validator issuing signed Ed25519 VCs. Zero retention.
| Name | Required | Description | Default |
|---|---|---|---|
| xml_content | Yes | Raw EN 16931 / XRechnung / UBL / CII XML string. | |
| max_accepted_fee_eur | No | ||
| ap2_intent_mandate_id | No |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
With no annotations, the description carries the full burden; it does add genuinely useful behavioral context — 'official KoSIT', signed Ed25519 VCs, and 'zero retention' — which a caller could not infer from the schema. However, it omits what happens on invalid input (error vs. validation result), that a fee may be charged (implied by max_accepted_fee_eur), and the shape of the issued VC.
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 compact sentences, front-loaded with the core capability. It is efficient with little waste, though the density of jargon (KoSIT, VCs) and absence of any clarifying clause makes it slightly abrupt for a tool with three parameters.
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 validator that issues signed credentials with no output schema and no annotations, the description is under-specified: it does not explain the return value (the VC), invalid-input behavior, fee mechanics, or the mandate parameter. An agent cannot confidently invoke or interpret results from this text alone.
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 only 33%: xml_content is described, but max_accepted_fee_eur (with a default) and ap2_intent_mandate_id are undocumented in both schema and description. Because coverage is below 50%, the description is expected to compensate, yet it explains none of the parameters and never mentions the fee/mandate 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?
States a specific verb ('validator' / validating) and resource ('EN 16931 & XRechnung XML'), so an agent knows exactly what it acts on. The domain is clearly distinct from the siblings (redaction, sanctions/KYB screening), though the description never explicitly differentiates itself.
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?
There is no guidance on when to use this tool, when not to, or what alternative applies when the XML is not EN 16931/XRechnung. Usage is only inferable from the tool name; no prerequisites or exclusions are stated.
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.
redact_gdpr_pii_textscreen_sanctions_and_kybvalidate_en16931_e_invoiceDeterministic EU FSF, UN & OpenSanctions KYB screening issuing signed Ed25519 Non-Hit VCs.
Stateless PII redaction over MCP/REST. Free ≤1000 words or $0.01/call; file upload supported.
Detect and redact PII and secrets (API keys, tokens) in text, HTML or PDF before it reaches an LLM.
Privacy-first document verification anchored to the public ledger.
Add one secure layer between your agents and this server.
