Skip to main content
Glama

Classify one AI system against the routes of the Regulation

classify_system
Read-only

Use this as the first step when you hold a concrete AI system and need to know which routes of the EU AI Act come into view: prohibited practice screening, Annex III high risk, transparency, FRIA and the general purpose model route, and with a version 2 profile the wider set of duties, such as Article 26 for a deployer. It runs the deterministic decision engine over coded answers and returns one applicability status per assessed obligation, with the timing and what is still open. It is not meant for a general question about the law. An obligation the engine did not assess is listed in assessment_scope.not_assessed with the reason; it was not found inapplicable. It does not return a single coded classification label.

Tool call template

Replace each placeholder with a value matching the input schema. Only required parameters are included; add optional parameters as needed.

{
  "id": 1,
  "jsonrpc": "2.0",
  "method": "tools/call",
  "params": {
    "arguments": {
      "profile": "<profile>"
    },
    "name": "classify_system"
  }
}

Input Schema

NameRequiredDescriptionDefault
langNoLanguage of the labels and explanations. Dutch is the default.nl
detailNobrief keeps the outcome, dates, reasons, open questions and the source with its locator; full adds the complete objects, statement texts and hashes, for re-computation or archiving.brief
profileYesThe coded answers about one system. The fifteen version 1 fields are required. A field that does not bear on this case takes not_relevant, and unsure is a real answer that leads to insufficient_context rather than to a guess. A version 1 profile has five obligations assessed. To have the wider set assessed, including Article 26 for a deployer of a high risk system, also send profile_version 2.0.0, chain_roles (consistent with actor_role) and the twelve other version 2 scope facts: ai_system_scope, jurisdiction_established_in_eu, jurisdiction_market_in_eu, jurisdiction_output_in_eu, scope_exclusion, annex_i_product, annex_i_third_party_conformity, article_25_role_change, provider_outside_eu, sandbox_participation, real_world_testing and adverse_significant_decision. The remaining version 2 facts are optional: an unanswered fact is unknown and leads to insufficient_context, never to a silent no. The backend rejects a profile whose answers contradict each other.
store_for_improvementNoOptional. Set to false when the user asks that Praxikon does not keep this request to improve its answers; only the anonymous daily count remains. See https://www.praxikon.com/nl/legal/privacy, section 2.7.

Output Schema

NameRequiredDescriptionDefault
viewNo
notesNo
datasetYes
privacyNo
profileNo
sourcesNo
summaryYes
disclaimerNo
provenanceYesWhich published endpoints produced this result. Every call is listed with its full URL, so a reader can re-issue it and check the answer rather than trust it.
obligationsYes
snapshot_idNo
blocking_flagsNo
deadline_notesNo
open_questionsNo
assessment_scopeNoWhat this assessment did and did not assess, from the engine itself. not_assessed groups the obligations that were not assessed by reason (for example profile_version_below_2), with the facts that would have them assessed. Not assessed is not the same as not applying.
evaluation_scopeNo
payload_hash_sha256No
decision_engine_versionNo
snapshot_schema_versionNo
evaluated_obligation_idsNo
not_indicated_obligation_idsNo

Schema Changelog

Changes observed during successful MCP inspections.

  1. Changed1 schema field changed
    • changedInput schema / properties / profile / properties / adverse_significant_decision / description
      Previous value: -"Can the output lead to a decision with legal or similarly significant adverse effects on a person (Article 86)?"New value: +"Does the deployer take a decision on the basis of the output from the high-risk AI system listed in Annex III (other than point 2) which produces legal effects for the affected person, or similarly significantly affects that person in a way that they consider to have an adverse impact on their health, safety or fundamental rights (Article 86(1))?"
  2. Changed36 schema fields changed
    • changedInput schema / properties / detail / description
      Previous value: -"Amount of data to return. brief keeps the outcome, the dates, the reasons, the open questions and the source with its locator; full adds the complete objects, every statement text and every hash, for re-computation or archiving."New value: +"brief keeps the outcome, dates, reasons, open questions and the source with its locator; full adds the complete objects, statement texts and hashes, for re-computation or archiving."
    • changedInput schema / properties / profile / properties / adverse_significant_decision / description
      Previous value: -"Version 2: can the output lead to a decision with legal or similarly significant adverse effects on a person (Article 86)?"New value: +"Can the output lead to a decision with legal or similarly significant adverse effects on a person (Article 86)?"
    • changedInput schema / properties / profile / properties / ai_system_scope / description
      Previous value: -"Version 2: is the assessed thing an AI system in the sense of Article 3(1)?"New value: +"Is the assessed thing an AI system in the sense of Article 3(1)?"
    • changedInput schema / properties / profile / properties / annex_i_product / description
      Previous value: -"Version 2: is the system a product, or a safety component of a product, under the Union legislation in Annex I?"New value: +"Is the system a product, or a safety component of a product, under the Union legislation in Annex I?"
    • changedInput schema / properties / profile / properties / annex_i_third_party_conformity / description
      Previous value: -"Version 2: does that product need a third party conformity assessment? not_relevant without an Annex I product."New value: +"Does that product need a third party conformity assessment? not_relevant without an Annex I product."
    • changedInput schema / properties / profile / properties / article_25_already_high_risk / description
      Previous value: -"Version 2, optional: with a role change, is the system concerned already high risk? not_relevant without a role change."New value: +"With a role change, is the system concerned already high risk? not_relevant without a role change."
    • changedInput schema / properties / profile / properties / article_25_role_change / description
      Previous value: -"Version 2: a role change under Article 25(1), or none."New value: +"A role change under Article 25(1), or none."
    • changedInput schema / properties / profile / properties / article_6_3_documented / description
      Previous value: -"Version 2, optional: is the Article 6(3) assessment documented?"New value: +"Is the Article 6(3) assessment documented?"
    • changedInput schema / properties / profile / properties / article_6_3_exception / description
      Previous value: -"Version 2, optional: which condition of Article 6(3) is invoked, if any."New value: +"Which condition of Article 6(3) is invoked, if any."
    • changedInput schema / properties / profile / properties / article_6_3_no_significant_risk / description
      Previous value: -"Version 2, optional: is there no significant risk of harm in the sense of Article 6(3)?"New value: +"Is there no significant risk of harm in the sense of Article 6(3)?"
    • changedInput schema / properties / profile / properties / article_6_3_profiling / description
      Previous value: -"Version 2, optional: does the system perform profiling of natural persons?"New value: +"Does the system perform profiling of natural persons?"
    • changedInput schema / properties / profile / properties / chain_roles / description
      Previous value: -"Version 2: the roles in the value chain, kept apart. Must agree with actor_role, for example [deployer] with actor_role deployer."New value: +"The roles in the value chain, kept apart. Must agree with actor_role, for example [deployer] with actor_role deployer."
    • changedInput schema / properties / profile / properties / complaint_to_authority / description
      Previous value: -"Version 2, optional: has a complaint been lodged with a market surveillance authority?"New value: +"Has a complaint been lodged with a market surveillance authority?"
    • changedInput schema / properties / profile / properties / controlled_input_data / description
      Previous value: -"Version 2, optional: does the deployer exercise control over the input data (Article 26(4))?"New value: +"Does the deployer exercise control over the input data (Article 26(4))?"
    • changedInput schema / properties / profile / properties / decisions_about_people / description
      Previous value: -"Version 2, optional: does the system make or assist decisions about natural persons (Article 26(11))?"New value: +"Does the system make or assist decisions about natural persons (Article 26(11))?"
    • changedInput schema / properties / profile / properties / gpai_commission_designated / description
      Previous value: -"Version 2, optional: has the Commission designated the model as having systemic risk?"New value: +"Has the Commission designated the model as having systemic risk?"
    • changedInput schema / properties / profile / properties / gpai_compute_above_threshold / description
      Previous value: -"Version 2, optional: is the training compute above the threshold of Article 51(2)?"New value: +"Is the training compute above the threshold of Article 51(2)?"
    • changedInput schema / properties / profile / properties / gpai_high_impact_capabilities / description
      Previous value: -"Version 2, optional: does the model have high impact capabilities under Article 51(1)(a)?"New value: +"Does the model have high impact capabilities under Article 51(1)(a)?"
    • changedInput schema / properties / profile / properties / gpai_systemic_notification / description
      Previous value: -"Version 2, optional: has the Commission been notified under Article 52(1)?"New value: +"Has the Commission been notified under Article 52(1)?"
    • changedInput schema / properties / profile / properties / jurisdiction_established_in_eu / description
      Previous value: -"Version 2: is the organisation established in the Union?"New value: +"Is the organisation established in the Union?"
    • changedInput schema / properties / profile / properties / jurisdiction_market_in_eu / description
      Previous value: -"Version 2: is the system placed on the market or put into service in the Union?"New value: +"Is the system placed on the market or put into service in the Union?"
    • changedInput schema / properties / profile / properties / jurisdiction_output_in_eu / description
      Previous value: -"Version 2: is the output used in the Union?"New value: +"Is the output used in the Union?"
    • changedInput schema / properties / profile / properties / law_enforcement_remote_identification / description
      Previous value: -"Version 2, optional: is the system used for post remote biometric identification in law enforcement (Article 26(10))?"New value: +"Is the system used for post remote biometric identification in law enforcement (Article 26(10))?"
    • changedInput schema / properties / profile / properties / logs_under_control / description
      Previous value: -"Version 2, optional: are the automatically generated logs under the control of the deployer (Article 26(6))?"New value: +"Are the automatically generated logs under the control of the deployer (Article 26(6))?"
    • changedInput schema / properties / profile / properties / personal_data_processing / description
      Previous value: -"Version 2, optional: does the use involve processing of personal data (Article 26(9))?"New value: +"Does the use involve processing of personal data (Article 26(9))?"
    • changedInput schema / properties / profile / properties / provider_outside_eu / description
      Previous value: -"Version 2: is the provider established outside the Union?"New value: +"Is the provider established outside the Union?"
    • changedInput schema / properties / profile / properties / public_authority_deployer / description
      Previous value: -"Version 2, optional: is the deployer a public authority or Union body (Article 26(8))?"New value: +"Is the deployer a public authority or Union body (Article 26(8))?"
    • changedInput schema / properties / profile / properties / real_world_testing / description
      Previous value: -"Version 2: is the system tested in real world conditions outside a sandbox?"New value: +"Is the system tested in real world conditions outside a sandbox?"
    • changedInput schema / properties / profile / properties / risk_detected / description
      Previous value: -"Version 2, optional: has a risk in the sense of Article 79(1) been detected?"New value: +"Has a risk in the sense of Article 79(1) been detected?"
    • changedInput schema / properties / profile / properties / safety_component / description
      Previous value: -"Version 2, optional: is the system used as a safety component?"New value: +"Is the system used as a safety component?"
    • changedInput schema / properties / profile / properties / sandbox_participation / description
      Previous value: -"Version 2: does the system take part in an AI regulatory sandbox?"New value: +"Does the system take part in an AI regulatory sandbox?"
    • changedInput schema / properties / profile / properties / scope_exclusion / description
      Previous value: -"Version 2: an exclusion under Article 2, or none."New value: +"An exclusion under Article 2, or none."
    • changedInput schema / properties / profile / properties / serious_incident / description
      Previous value: -"Version 2, optional: has a serious incident occurred?"New value: +"Has a serious incident occurred?"
    • changedInput schema / properties / profile / properties / third_party_components_integrated / description
      Previous value: -"Version 2, optional: does a third party supply components that are integrated in the system?"New value: +"Does a third party supply components that are integrated in the system?"
    • changedInput schema / properties / profile / properties / workplace_use / description
      Previous value: -"Version 2, optional: is the system put into service or used at the workplace (Article 26(7))?"New value: +"Is the system put into service or used at the workplace (Article 26(7))?"
    • changedInput schema / properties / store_for_improvement / description
      Previous value: -"Optional. Set to false when the user asks that Praxikon does not keep this request to improve its answers; nothing of this call is then kept beyond the anonymous daily count. See https://www.praxikon.com/nl/legal/privacy, section 2.7."New value: +"Optional. Set to false when the user asks that Praxikon does not keep this request to improve its answers; only the anonymous daily count remains. See https://www.praxikon.com/nl/legal/privacy, section 2.7."
  3. Changed1 schema field changed
    • addedInput schema / properties / store_for_improvement
      Added value: +{
      +  "default": true,
      +  "description": "Optional. Set to false when the user asks that Praxikon does not keep this request to improve its answers; nothing of this call is then kept beyond the anonymous daily count. See https://www.praxikon.com/nl/legal/privacy, section 2.7.",
      +  "type": "boolean"
      +}
  4. First observed

TDQS

A4.5/5.0
Behavior5/5

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

Beyond the readOnlyHint, openWorldHint and destructiveHint annotations, the description explains that a deterministic decision engine runs over coded answers, returns one applicability status per assessed obligation with timing and open points, never returns a single coded label, and lists unassessed obligations in assessment_scope.not_assessed with reasons rather than treating them as inapplicable. This adds substantial behavioral context about how to interpret the output and its limits.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness5/5

Is the description appropriately sized, front-loaded, and free of redundancy?

The description is front-loaded with the primary use case and then adds only relevant behavioral caveats. Every sentence contributes to purpose, scope, output interpretation or exclusions, with no filler.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness5/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

For a complex classification tool with a large nested profile schema, an output schema and read-only annotations, the description covers the purpose, the route set, the version 1 versus version 2 distinction, the deterministic nature of the engine and the meaning of unassessed obligations. Nothing essential for correct selection and interpretation is missing.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters3/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

The input schema has 100% description coverage and documents the profile fields in detail, so the schema already carries the parameter semantics. The description adds only high-level context about coded answers and version 2 profiles; it does not meaningfully explain individual parameters beyond what the schema provides.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose5/5

Does the description clearly state what the tool does and how it differs from similar tools?

The description states a specific verb and resource: classify one concrete AI system against the routes of the EU AI Act, naming the prohibited-practice, Annex III, transparency, FRIA and GP model routes. It also distinguishes the tool from a general legal question and from a single coded classification label, so an agent can tell what it is for.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines4/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

It gives a clear when-to-use condition as the first step for a concrete AI system, and an explicit when-not-to-use condition for general questions about the law. It does not name alternative sibling tools such as answer_question or get_applicable_obligations, so the routing guidance is clear but not fully explicit about alternatives.

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

Try in Browser

Glama MCP Gateway

Add one secure layer between your agents and this server.

Resources