Products¶
Five Products. One Agentic Positioning Engine. ¶
The CRP ecosystem positions every agent on the right task, with the right context and tools, then produces signed, tamper-evident proof that your safety, grounding, and policy controls operated. EU AI Act, AIUC-1, ISO 42001, NIST AI RMF, or SOC 2-for-AI: the same runtime data becomes audit-ready evidence.
Context Relay Protocol™ powers production-ready products built on top of the core protocol. Every product leverages CRP's 6-stage extraction, HMAC audit trail, and continuation engine - giving you capabilities that no other framework provides.
CRP Gateway and CRP Comply are on the managed-cloud waitlist. All products can also be self-hosted. White-label and OEM deployments are available for Enterprise.
CRP Gateway¶
Managed-cloud waitlist
The AI Safety Firewall - govern every LLM call with one base_url change.
A drop-in HTTPS endpoint that adds the Decision Provenance Engine, Safety Policy enforcement, automatic continuation, and HMAC audit emission to any OpenAI-compatible LLM call. Non-bypassable safety for every service that points its LLM client at the gateway.
- Best for: Teams that want safety, continuation, and audit on every LLM call without rebuilding their app.
- Key outcome: Every call returns risk, grounding, and provenance; long tasks finish instead of truncating.
- DPE pipeline (SPEC-005) on every response
CRP-Safety-Policyenforcement at the transport layer- Audit chain emission into CRP Comply automatically
- AIUC-1: Security, Safety, Reliability controls
- Managed, self-hosted, sidecar, or edge deployment
- Level 2 conformant (SPEC-014)
- Developer $29/mo · Team $199/mo
CRP Comply¶
Managed-cloud waitlist
The Compliance Automation Engine - turn runtime evidence into regulator-ready compliance.
A compliance gateway + dashboard that routes every LLM call through 13+ CRP security subsystems and generates EU AI Act, ISO 42001, GDPR, and AIUC-1 evidence packs from your actual system behaviour - not consultant PDFs.
- Best for: Organisations facing EU AI Act, ISO 42001, GDPR, or AIUC-1 deadlines that need audit-ready evidence.
- Key outcome: DPIAs, technical docs, and conformity packs generated in seconds from cryptographic audit trails.
- OpenAI-compatible proxy with PII scanning & injection detection
- HMAC-signed tamper-evident audit trail (7 regulatory frameworks)
- GDPR Art. 17 erasure, Art. 30 processing records, consent management
- AIUC-1: Accountability, Society, Data & Privacy controls
- One-click evidence packs for regulators
- Free · Starter $49/mo · Scale $499/mo · Enterprise
CRP Scan¶
Available
The Pre-Production Security Scanner - catch ungoverned AI calls before they reach production.
Add one workflow step and every pull request is checked for AI-system patterns that need CRP coverage: unaudited LLM calls, missing safety policies, ungoverned context construction, and regulatory mapping gaps. Free for public repos.
- Best for: Engineering teams that want governance gaps caught in CI, not during an audit.
- Key outcome: HIGH/CRITICAL findings are blocked before merge; auto-remediation PRs fix routing and policy stubs.
- Catches HIGH/CRITICAL governance gaps in PRs
- SARIF upload to GitHub code scanning
- Auto-remediation PRs with Gateway routing + policy stubs
- AIUC-1: Security, Accountability controls
- Specified in SPEC-013
CRP Scribe¶
Roadmap
Unlimited Long-Form Document & Book Generation
A generation engine that uses 28+ CRP subsystems to produce complete multi-chapter documents without hitting the LLM output wall. CRP's continuation engine, knowledge graph, voice profiling, and 6-stage extraction pipeline carry context across chapters - producing 50K+ word documents that are coherent from start to finish.
- 5 built-in templates (whitepaper, manual, guide, report, book)
- Per-chapter quality scoring (S/A/B/C/D tiers)
- Works with OpenAI, Anthropic, Ollama, LM Studio
CRP Visualise¶
Roadmap
Live Session Tracing for Auditors
Renders the Window DAG, DPE verdicts, claim provenance, hallucination heat-maps, and Safety Policy evaluations for any CRP session - in real time, in the browser. Sealed read-only mode for external auditors and regulators.
- Window DAG explorer + continuation timeline
- Claim-to-fact provenance with click-through
- Safety Policy evaluation trail
- HMAC chain integrity verification
Compare at a glance¶
| Product | What it does | Best for | Starting price |
|---|---|---|---|
| CRP Gateway | OpenAI-compatible proxy with DPE, safety policy, continuation, and audit | Teams adding governance to every LLM call | $29/mo Developer |
| CRP Comply | Compliance dashboard + evidence pack generator from runtime audit data | Organisations facing EU AI Act / ISO 42001 / GDPR deadlines | Free (100 calls/mo) |
| CRP Scan | GitHub Action + CLI scanner for ungoverned AI calls and remediation PRs | Teams that want CI governance gates | Free for public repos |
| CRP Visualise | Live session DAG, provenance, and risk browser (roadmap) | Auditors and incident responders | Roadmap / waitlist |
| CRP Scribe | Long-form document & book generation engine (roadmap) | Authors, analysts, and report teams | Roadmap / waitlist |
One SDK for every product¶
All products are reachable through the same CRP Python SDK:
import crp
client = crp.SDKClient(provider="openai", model="gpt-4o")
client.ingest("./docs/")
answer = client.ask("Write a compliance report", depth="thorough")
print(answer.text)
print(answer.quality) # S/A/B/C/D
print(answer.sources) # cited facts
print(answer.crp.risk) # LOW/MEDIUM/HIGH/CRITICAL
Coming Soon¶
CRP Tutor¶
Roadmap
Adaptive Learning & Training Platform
Multi-stage training programs with knowledge assessment, gap analysis, and personalized learning paths - all backed by CRP's knowledge fabric.
CRP Analyst¶
Roadmap
Strategic Analysis Engine
Cross-document analysis, trend identification, and insight extraction across large document corpora with full provenance tracking.
CRP Code¶
Roadmap
Context-Aware Code Generation
Multi-file code generation with architecture awareness, test scaffolding, and documentation - maintaining coherence across arbitrarily large codebases.
All Products Include¶
| Feature | Description |
|---|---|
| CRP Core | 9 dispatch strategies, 6-stage extraction, continuation engine |
| HMAC Audit Trail | Cryptographically signed session evidence |
| Quality Assessment | S/A/B/C/D tiers with 4-signal completion detection |
| Provider Agnostic | OpenAI, Anthropic, Ollama, LM Studio - your choice |
| Elastic License 2.0 | Free for your applications, protected from service theft |
Contact¶
-
General Enquiries
-
Enterprise & Licensing