Skip to main content

Python SDK for CRP-Comply — EU AI Act & GDPR compliance automation

Project description

crp-comply-sdk

PyPI Python License

Python SDK for CRP-Comply — EU AI Act & GDPR compliance automation. A thin, typed HTTP client covering every user-facing endpoint of the CRP-Comply REST API.

Install

pip install crp-comply-sdk

Quick Start

from crp_comply_sdk import CRPComply

with CRPComply(api_key="crp_...") as client:
    # Account
    me = client.me()
    usage = client.usage()

    # EU AI Act Article 6 risk classification
    risk = client.risk_assessment(
        system_name="Hiring Assistant",
        category="employment",
        affects_fundamental_rights=True,
    )

    # GDPR Article 35 DPIA
    dpia = client.dpia(
        system_name="Hiring Assistant",
        data_subjects=["candidates", "employees"],
        makes_automated_decisions=True,
    )

    # Full conformity evidence pack (zip with manifest + HMAC)
    pack = client.evidence_pack(system_name="Hiring Assistant", category="employment")
    zip_bytes = client.download_evidence_pack(pack["pack_id"])
    with open("evidence.zip", "wb") as f:
        f.write(zip_bytes)

API Surface

All methods are keyword-only. Every call is tier-gated server-side — a CRPComplyTierError is raised with an upgrade_url when a feature isn't available on your plan.

Compliance generation

Method Endpoint Purpose
risk_assessment(...) POST /risk-assessment EU AI Act Article 6
compliance_report(..., markdown=False) POST /compliance-report[/markdown] Full compliance status
dpia(...) POST /dpia GDPR Article 35 DPIA
transparency(system_name=...) POST /transparency AI Act Article 13
technical_docs(system_name=...) POST /technical-docs AI Act Article 11
audit_session(session_file=...) POST /audit Audit a CRP session
full_report(...) POST /full-report Markdown mega-report

Evidence packs

Method Endpoint
evidence_pack(...) POST /evidence-pack — build new pack
list_evidence_packs() GET /evidence-packs
get_evidence_pack(pack_id) GET /evidence-packs/{id} — manifest
download_evidence_pack(pack_id) GET /evidence-packs/{id}/download — raw zip bytes
delete_evidence_pack(pack_id) DELETE /evidence-packs/{id}

Persisted reports

Method Endpoint
list_reports(kind=None) GET /reports?kind=...
get_report(report_id) GET /reports/{id}
get_report_markdown(report_id) GET /reports/{id}/markdown — returns str
delete_report(report_id) DELETE /reports/{id}

SDK gateway (realtime checks)

Method Endpoint
features() GET /sdk/features — tier feature matrix
audit(prompt=..., response=...) POST /sdk/audit — PII + injection + risk
classify_risk(...) POST /sdk/classify — quick AI Act bucket

Account

Method Endpoint
health() GET /health (unauthenticated)
me() GET /me — profile + tier + provider status
usage() GET /usage — monthly quota breakdown

Configuration

Option Env var Default
api_key CRP_COMPLY_API_KEY — (required)
base_url CRP_COMPLY_BASE_URL https://comply.crprotocol.io/api/v1
timeout 60 seconds
import os
os.environ["CRP_COMPLY_API_KEY"] = "crp_..."
os.environ["CRP_COMPLY_BASE_URL"] = "https://comply.crprotocol.io/api/v1"

from crp_comply_sdk import CRPComply
client = CRPComply()  # picks up env vars

LLM Backends

CRP-Comply is LLM-agnostic. Configure your preferred backend server-side at Settings → LLM Provider. Supported:

Backend Default URL Notes
OpenAI https://api.openai.com/v1 Cloud
Anthropic https://api.anthropic.com/v1 Cloud
LM Studio http://localhost:1234/v1 Local, OpenAI-compatible
Ollama http://localhost:11434/v1 Local, OpenAI-compatible
Custom any Any OpenAI-compatible endpoint

Exception Hierarchy

CRPComplyError              # base — all SDK errors
├── CRPComplyAuthError      # 401, 403 — invalid / missing API key
├── CRPComplyQuotaError     # 429 — monthly quota exhausted (has .upgrade_url)
├── CRPComplyTierError      # 402 — feature not in tier (has .upgrade_url,
│                           #       .feature, .current_tier, .required_tier)
└── CRPComplyServerError    # 5xx — server-side failure
from crp_comply_sdk import CRPComply, CRPComplyTierError, CRPComplyQuotaError

try:
    pack = client.evidence_pack(system_name="X", category="y")
except CRPComplyTierError as exc:
    print(f"Upgrade required: {exc.upgrade_url}")
except CRPComplyQuotaError as exc:
    print(f"Quota exhausted. Upgrade at {exc.upgrade_url}")

License

Apache-2.0 — see LICENSE.

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

crp_comply_sdk-0.2.0.tar.gz (16.0 kB view details)

Uploaded Source

Built Distribution

If you're not sure about the file name format, learn more about wheel file names.

crp_comply_sdk-0.2.0-py3-none-any.whl (13.9 kB view details)

Uploaded Python 3

File details

Details for the file crp_comply_sdk-0.2.0.tar.gz.

File metadata

  • Download URL: crp_comply_sdk-0.2.0.tar.gz
  • Upload date:
  • Size: 16.0 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/6.2.0 CPython/3.13.12

File hashes

Hashes for crp_comply_sdk-0.2.0.tar.gz
Algorithm Hash digest
SHA256 f76580c68af6d940bcb923d7819fe3675c15d376ac3638512f3946ed29fb5063
MD5 549c69f4769078076f4878b4cf322493
BLAKE2b-256 9fc22802aa13a0778c109219950ccd6daafdd5c4ae71fce800cbb32ddee5b1ad

See more details on using hashes here.

File details

Details for the file crp_comply_sdk-0.2.0-py3-none-any.whl.

File metadata

  • Download URL: crp_comply_sdk-0.2.0-py3-none-any.whl
  • Upload date:
  • Size: 13.9 kB
  • Tags: Python 3
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/6.2.0 CPython/3.13.12

File hashes

Hashes for crp_comply_sdk-0.2.0-py3-none-any.whl
Algorithm Hash digest
SHA256 fdab4ba02b11347f5a835f7e6e78e243301e2994fcebaa62039fc7e348735487
MD5 fee55d9dd6d713a0ba092fd97c593b51
BLAKE2b-256 9aa19fe30a62eb6829ef7476df31c784ddc90de36609a2cb2c76f47972966e1c

See more details on using hashes here.

Supported by

AWS Cloud computing and Security Sponsor Datadog Monitoring Depot Continuous Integration Fastly CDN Google Download Analytics Pingdom Monitoring Sentry Error logging StatusPage Status page