Eu Ai Compliance

EU AI Act risk classification, Article 12 compliance logging, and self-assessment reports — deadline August 2, 2026

Security Scan

The skill is an instruction-only wrapper that directs the agent to a remote EU AI Act compliance API; its declared requirements and runtime instructions are generally consistent with the stated purpose, with only minor documentation/clarity issues to verify before use.

Purpose & Capability

Name/description match the SKILL.md: it provides risk classification, Article 12 mapping, and paid reports via HTTP endpoints. The skill requests no local binaries, env vars, or installs — which is consistent for an instruction-only HTTP API integration. Minor inconsistency: the SKILL.md metadata homepage (soulledger.sputnikx.xyz) differs from the base API (soul.sputnikx.xyz) and the ownership/hosting is 'unknown', so verify provider identity before relying on results.

Instruction Scope

Instructions only tell the agent to curl specific API endpoints on the listed domain — there are no commands to read local files or other system state. However, the skill gives no guidance about what 'agent_id' is or about authentication/payment flows. Also: any data you send (descriptions, system details) will go to an external service — avoid sending secrets or proprietary data until you've validated the provider and privacy terms.

Install Mechanism

No install spec or code is present (instruction-only). This minimizes on-disk risk because nothing is downloaded or executed locally by the skill itself.

Credentials

No environment variables, credentials, or config paths are requested. That is proportionate for a client that simply calls a public web API. Note: the SKILL.md mentions payments and on-chain anchoring but does not request wallet keys — payment likely occurs externally; confirm how payments are processed before sending funds.

Persistence & Privilege

always is false and model invocation is allowed (default). The skill does not request persistent privileges, nor does it modify system or other-skill configuration — this is appropriate for the described function.

Assessment

This skill is an instruction-only connector to an external compliance service; that makes it low local risk but means you are trusting a third party with any text you send. Before installing or using it: 1) Verify the provider and website (the SKILL.md lists two related but different domains). 2) Do not send sensitive or proprietary system details until you confirm the service's privacy policy and data retention practices. 3) Clarify the payment flow and costs (the pricing string in SKILL.md is ambiguous: 'x402 USDC'); avoid sending funds until you understand how payments are handled. 4) If you need cryptographic proofs (Merkle root anchoring on Base chain), ask for verifiable examples and public anchors you can independently check. 5) For production use, run initial tests with non-sensitive sample data.

EU AI Act Compliance

Free risk classification + paid full assessments. EU AI Act enforcement deadline: August 2, 2026. Fines: EUR 35M or 7% global turnover.

Base URL

https://soul.sputnikx.xyz

Endpoints

Risk Classification (FREE)

curl "https://soul.sputnikx.xyz/soul/compliance/risk-classification?description=facial+recognition+for+hiring"

Returns: { risk_level, reasoning, annex_iii_matches, disclaimer } — instant HIGH/LIMITED/MINIMAL classification.

Article 12 Mapping (FREE)

curl https://soul.sputnikx.xyz/soul/compliance/mapping

Returns: Full Article 12 → data field mapping for compliance logging requirements.

Self-Assessment ($1.00 x402 USDC)

curl https://soul.sputnikx.xyz/soul/compliance/self-assessment/{agent_id}

Complete EU AI Act self-assessment report with recommendations.

Annex IV Technical Documentation ($1.00 x402 USDC)

curl https://soul.sputnikx.xyz/soul/compliance/annex-iv/{agent_id}

Annex V EU Declaration ($0.50 x402 USDC)

curl https://soul.sputnikx.xyz/soul/compliance/annex-v

Full Compliance Report ($2.00 x402 USDC)

curl https://soul.sputnikx.xyz/soul/compliance/full-report

Complete bundle: risk classification + self-assessment + Annex IV + Annex V.

Web Interface

Free compliance checker: https://soulledger.sputnikx.xyz/compliance

Infrastructure

• Hash-chain logging (SHA-256, append-only)
• Merkle root anchoring on Base chain
• Behavioral drift detection
• Runtime monitoring (not one-time reports)

When to use this skill

• Check if your AI system is EU AI Act compliant
• Generate compliance documentation before the August 2026 deadline
• Implement Article 12 automatic logging
• Get risk classification for any AI system description