بسم الله الرحمن الرحيم

واجهة برمجة التطبيقات

IPSC API — three tiers, audit-grade, on Azure

IPSC is exposed as a commercial HTTP API on Microsoft Azure infrastructure with physically separate indexes per access tier, semantic-search over an 11-layer vector embedding stack, and provenance stamped on every response. Public, Research, and Scholar tiers are issued under separate licensing terms.

v3.4 deployed · commercial access via licensing
المستويات الثلاثة Three access tiers

Tier separation is the central design decision of the IPSC API. Each tier is a physically separate index — the Public surface does not contain Scholar-tier fields at all, so field-exposure cannot be accidentally widened by a query parameter or a misconfigured client.

Public

Audience
App developers, casual researchers, public-facing Islamic-content products
Licensing
Attribution-only license, throttled rate limit
What this tier surfaces
Core hadith record, computed grade, attestation context, Arabic source text, English translation, and semantic-search over the matn vector layer.

Research

Audience
Academic researchers, university Islamic-studies programs, app teams building grading or chain-analysis features
Licensing
Paid commercial subscription
What this tier surfaces
Adds parsed isnād chain (with narrator IDs and tier labels), chain-quality index, classical-defect dimensions (shudhūdh, idrāj, maqlūb, mudtarib), ʿilal cross-references, and the full v3.26 enrichment provenance flags.

Scholar

Audience
Peer reviewers, qualified muḥaddithīn, scholarly-review institutions
Licensing
Identity-verified reviewer agreement, no redistribution
What this tier surfaces
Adds full multi-pass matn-criticism reasoning (with classical citation), per-narrator primary classical assessment, defense-document join, classical-dimension detection metadata, and the NRS audit-flag layer.
البنية التحتية Infrastructure & credibility

The corpus is served from production-grade Azure infrastructure with regression testing on every release, atomic version cutover, and a non-strippable attribution block on every response.

13 indexes

Tier-separated indexes

Public, Research, and Scholar surfaces are physically separate indexes, not row-level filtered views — defense-in-depth on field exposure.

11 vector layers

Semantic-search coverage

OpenAI text-embedding-3-large at 3,072 dimensions, unit-normalized, across matn, narrators, defects, and glossary terms.

v3.4 deployed

Production version

IPSC v3.4 currently live; v3.26 staged for next deployment window. Atomic version cutover supported.

Provenance

Stamped per response

Every API response carries corpus version, hash, tier scope, and the canonical attribution block — non-strippable by license.

طلب الوصول Commercial access

Endpoint specifications, request headers, response envelope, rate limits, and the Field Exposure Policy are issued to licensed customers as part of onboarding. Choose the tier that matches your use case and reach out for licensing terms and credentials.

Public tier App developers · attribution license Research tier Academic & commercial · paid subscription Scholar tier Identity-verified reviewer agreement

For the methodology behind the corpus, see /methodology. For the audit-and-quality cycle, see /honesty. For the full provenance disclosure that ships with every record, see /provenance.