Skip to main content
Glama
Sign In

Privi

by com.onrender.privi-ledger
OverviewDiscussionsAdmin

AI Act: register an AI system

ai_act_system_save
Idempotent

Add or update an AI system in your company's EU AI Act register. Privi screens it against the Article 5 bans, classifies it (prohibited / high-risk / transparency / general-purpose / minimal), and lists the obligations that apply to your role with their dates. Every change is versioned and written to your tamper-evident 'ai-act' log. Requires the organization API key.

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": {
      "system": "<system>",
      "actor": "<actor>"
    },
    "name": "ai_act_system_save"
  }
}

Input Schema

NameRequiredDescriptionDefault
actorYesWho is making the change (person or agent)
mergeNoMerge into the existing record (true) or replace it (false)
systemYesAI system record. Required: system_id. Recommended: name, owner, role (provider|deployer|provider_and_deployer|importer|distributor), intended_purpose, used_in_eu, vendor_name. Risk flags: annex_iii_area (biometrics|critical_infrastructure|education|employment|essential_services|law_enforcement|migration|justice_democracy or empty), annex_i_product, gpai_model, public_body_or_public_service, interacts_with_people, generates_synthetic_content, emotion_or_biometric_categorisation, deepfake_or_public_interest_text, not_high_risk_justification (Art. 6(3)). prohibited_screen {practice: true/false} for: harmful_manipulation, exploiting_vulnerabilities, social_scoring, predictive_policing_profiling, untargeted_face_scraping, workplace_school_emotion_recognition, biometric_categorisation_sensitive, realtime_remote_biometric_id_public, non_consensual_intimate_imagery. controls {obligation_id: {done, evidence, date}} for: ai_literacy, disclose_ai_interaction, mark_synthetic_content, inform_emotion_biometric, disclose_deepfake, gpai_technical_documentation, gpai_copyright_policy, gpai_training_summary, risk_management, data_governance, technical_documentation, automatic_logging, instructions_for_use, human_oversight_design, accuracy_robustness_security, quality_management, conformity_assessment, eu_database_registration, post_market_monitoring, use_per_instructions, human_oversight_assigned, monitor_and_report, keep_logs_6_months, inform_workers, inform_affected_persons, fundamental_rights_impact_assessment. log_streams: Privi Ledger stream IDs holding the system's logs. retired: true to retire.

Schema Changelog

Changes observed during successful MCP inspections.

  1. Added

TDQS

A4/5.0
Behavior5/5

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

Beyond the annotations (which only say non-readOnly, idempotent, non-destructive), the description discloses substantial behavior: it screens against Article 5 bans, auto-classifies risk tier, enumerates applicable obligations with dates, versions every change, and writes to a tamper-evident 'ai-act' log. It also states the org API key requirement, which is exactly the kind of auth context annotations cannot convey.

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?

Four compact sentences, each adding distinct information (purpose, processing pipeline, audit trail, auth requirement), with no repetition or filler. The core action is front-loaded in the first sentence.

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

Completeness4/5

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

For a save tool with no output schema, the description adequately explains what the operation does downstream (screening, classification, obligations, versioned log write). It does not describe the response shape or how classification results are surfaced back to the caller, which is a minor gap given no output schema exists.

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?

Schema description coverage is 100%, and the nested 'system' property already documents an extensive field list, so the schema carries the parameter burden. The description adds no syntax or format detail about system, actor, or merge beyond what the schema states. Baseline 3 is appropriate when the schema does the heavy lifting.

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

Purpose4/5

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

The description states a specific verb pair (add/update) and a precise resource: an AI system in the company's EU AI Act register. It goes further by naming the processing that follows (Article 5 screening, classification, obligation listing), which makes the tool's scope vivid. It stops short of explicitly distinguishing itself from close siblings like ai_act_register or ai_act_check, so it is clear but not fully differentiated.

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

Usage Guidelines3/5

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

Usage is implied by 'add or update' and the org API key prerequisite, but the description never says when to choose this over ai_act_check, ai_act_register, or ai_act_export, nor any when-not conditions. The auth requirement is the only concrete calling precondition given.

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

All Privi toolsInspect
Try in Browser
One endpoint for every MCP toolRayrun runs every MCP server behind one URL and holds the credentials. Each call is decided per client, tool, and argument, then recorded.ray.runAd

Glama MCP Gateway

Add one secure layer between your agents and this server.

Resources