Cortex AI

Cortex is Kliper’s built-in AI assistant. It is a full agentic loop with access to 15 tools, grounded retrieval from your firm’s knowledge base, PII redaction, and built-in safety checks.

At a high level, Cortex can:

• Read — pull requirement details, assessment answers, evidence files, scoping data, and existing tasks
• Write — draft and save justifications, testing procedure responses, tasks, and calendar events directly back to the assessment
• Search — retrieve grounded context from your firm’s past ROCs, accepted evidence, and PCI DSS guidance
• Reason — iterate across multiple tool calls in a single turn, deciding its next step based on intermediate results
• Review — flag findings that may not hold up, score each flag with CRESS reliability, draft closure plans, and triage everything across engagements in one queue (see below)

All features accelerate the assessment process without replacing assessor judgment. Cortex produces drafts; the assessor retains full authority over all final content.

Plans & Allowances

Cortex chat is included on every plan, including Free — no feature gate, just a monthly message allowance. Write operations meter separately, so browsing and asking questions never burns your drafting budget.

The AI-operations allowance covers two independently tracked meters:

• AI drafts — Cortex-chat write tools (saving justifications, TP responses, tasks) and the Section 7 justification drafter.
• File summaries — AI file analysis and diagram extraction. The first AI analysis of each file is free on every plan; only re-analyses meter.

Read-only chat messages count against the message allowance only. When an allowance runs out, Cortex tells you which plan raises it — existing drafts and analyses are never locked away. Live meters for all three allowances are on the Billing page.

The Agent Loop

Cortex is not a single-shot chat model. Every user message kicks off an iterative loop:

Available Tools

Cortex has 16 tools grouped by purpose:

Read tools

Write tools

Search tools

Draft from Evidence

In the Section 7 requirement editor, Cortex can draft a finding justification directly from the evidence you’ve already cited — human-in-the-loop end to end:

Each draft consumes one AI draft from the monthly allowance.

Structured citations

When Cortex answers from assessment content, replies carry structured citations — the specific reporting-instruction items behind an answer and links to the interviewees involved — so an assessor can verify a claim at its source instead of taking prose on faith.

Cortex reads the whole ROC, not just Section 7

For most of Cortex’s life it could only see the Section 7 requirement tables. Ask it about the Executive Summary and it answered with boilerplate — not because it declined, but because that content genuinely never reached it. On a real assessment that meant the majority of the saved document — the Executive Summary, all of Parts I and II, the quarterly scan results, sampling, and the appendices — was invisible to it.

Cortex now reads those report sections too, through get_report_section. A few things follow from that:

• Ask about the Executive Summary, appendices, sampling, or scan results by name and Cortex draws on what is actually saved in those sections, the same as it does for a requirement.
• Names, not numbers, come back. Cortex refers to “6.4 Documentation Evidence,” not “section 6.4,” so a report section is never confused with the requirement that shares its number.
• Empty is reported honestly. A section you have not filled in yet is described as empty, with a note on what it expects — never as filled, and never as a raw blank value.

Behavioral Rules

Cortex’s system prompt enforces a strict set of behaviors that cannot be bypassed by user messages:

Status Indicators

Every Cortex response shows a live, descriptive status while the agent runs:

The old generic “Cortex is thinking…” has been replaced across the board.

Your conversation is kept

Closing the Cortex panel never ends the conversation. Reopen it — even after a page refresh or a deploy that reloads the tab — and the thread is still there, along with any prompt you had typed but not sent yet.

Within a thread, Cortex works from what you have actually been discussing. It loads the most recent messages in the conversation, so a follow-up like “and the one after it” or “fix that section” is read against your last few turns — not anchored back to the first thing you asked when the thread began.

Knowledge Base & RAG

Cortex can ground its drafts in your firm’s past work. The Knowledge Base is a dedicated ingestion pipeline that turns ROCs, AOCs, meeting transcripts, and accepted evidence into retrievable context.

How Ingestion Works

How Cortex Uses It

When drafting a justification or TP response, Cortex can call the search_firm_knowledge tool with a semantic query (e.g., “audit log retention policy 12 months”) scoped to the current requirement. The tool returns top-K relevant chunks with metadata (source document, requirement, upload date) which Cortex then references in its draft.

Admin UI

The Knowledge Base panel (admin-only) shows:

Jobs poll every 3 seconds while any ingestion is active, so status updates are near-real-time.

Evidence Validation

What It Does

When an evidence file is uploaded and tagged to a specific PCI DSS requirement, Cortex can validate whether the document adequately covers the content items that the ROC template requires for that requirement.

Kliper maintains a validation specification for each requirement — a structured checklist of content items the evidence document must address. These specs are derived from the PCI DSS v4.0.1 ROC template and stored in document-validation.json.

Validation Flow

{
  "requirementId": "3.4.1",
  "title": "PAN rendering requirement",
  "type": "document",
  "tag": "DOCFW",
  "checkedAt": "2026-02-28T14:30:00.000Z",
  "items": [
    {
      "criterion": "Document defines encryption algorithms used for PAN storage",
      "found": true,
      "excerpt": "AES-256 encryption is applied to all PAN data at rest...",
      "note": null
    },
    {
      "criterion": "Document specifies key management procedures",
      "found": false,
      "excerpt": null,
      "note": "No key management section found in document"
    }
  ],
  "summary": {
    "total": 8,
    "found": 6,
    "missing": 2,
    "status": "partial"
  },
  "model": "gpt-4o-mini",
  "tokensUsed": { "input": 4200, "output": 850 }
}

Validation Statuses

The summary status is derived from the found/total ratio:

What the Assessor Sees

In the Attachments Panel, each file displays its validation status. Expanding the validation result shows:

• A checklist of all criteria with checkmarks (found) or X marks (not found).
• Excerpts from the document that demonstrate coverage.
• Notes on partial coverage or missing items.
• The AI model used and when the validation was performed.

Inline analysis & requirement matching

When you open a file’s analysis, Cortex shows it inline (not just a pass/fail checklist): a relevance assessment, a short summary, suggested tags, and a requirement-match grid with confidence bars showing which PCI DSS requirements the file appears to support — with an “Analyzed by Cortex” footer. While it runs, a live analyzing pipeline shows the stages: Security scan → Extracting → Summarizing → Matching.

Cortex Autofill — ROC Findings Generation

What It Does

Cortex Autofill generates a draft findings description for a specific PCI DSS requirement. This is the narrative text that appears in the final ROC, describing what the assessor examined, what methods were used, and what was observed.

When to Use It

Autofill is most effective when the assessor has already:

1. Uploaded relevant evidence files and tagged them to the requirement.
2. Filled in at least some testing procedure responses.
3. Selected a finding status (In Place, Not Applicable, Not Tested, Not in Place).

Cortex will work with incomplete data, but it will flag what is missing and use placeholders ([PENDING_RESPONSE]) rather than fabricating content.

How It Works

{
  "content": "Per DOCFW, firewall rulesets restrict inbound traffic to required ports and protocols only. Configuration screenshots in EVDFW show deny-all default rules on external-facing interfaces. Network administrator interview confirmed change management procedures are followed for all modifications.",
  "warnings": [
    "Assessor responses missing for: 1.2.3.b, 1.2.3.c",
    "No evidence files uploaded for Requirement 1."
  ]
}

Autofill with Compensating Controls

When the assessor selects the Compensating Control method, Cortex adjusts its output to note that Appendix C applies and frames the findings around the compensating control rather than the standard testing procedure.

Autofill with Customized Approach

When the assessor selects the Customized Approach method, Cortex:

1. Loads the Customized Approach Objective from PCI DSS guidance for that requirement.
2. Instructs the AI to explain how the entity’s implementation meets the Customized Approach Objective, rather than addressing the standard defined approach testing procedures.
3. If no Customized Approach Objective exists for the requirement (some requirements are not eligible), a warning is returned.

Validation Step Analysis

Cortex also analyzes the reporting instruction text to determine which validation steps are relevant for a requirement. It uses keyword matching to identify required evidence types:

An Assessor step is always included regardless of keywords.

These steps populate the structured prefix section of the requirement answer, ensuring that the ROC includes complete documentation of what was examined.

Cortex Review and Triage

Beyond drafting, Cortex acts as an advisory reviewer of your completed work — flagging findings that may not hold up, scoring how much to trust each flag, and rolling everything into a cross-engagement queue. It is advisory only: Cortex never changes your deterministic gap/risk scores or your answer data, and the layer is gated per workspace (active only when Cortex is enabled).

CRESS reliability scoring

Instead of a model self-reporting “90% confident,” Cortex scores each verdict with CRESS — a reliability score (0–100) computed from five weighted signals: self-consistency across samples (30%), grounding (25% — did the verdict cite and quote a specific missing element), input sufficiency (20%), the model’s self-reported confidence (15%), and requirement difficulty (10%). The score maps to a band:

CRESS replaces self-reported confidence wherever Cortex surfaces a verdict.

Advisory finding-review layer

On the Risk and Gap views, Cortex marks each reviewed control as addressed / partial / not addressed, with stale-detection when the underlying answer changes after a review. These verdicts sit alongside — never replace — your deterministic scores.

• Dismiss a verdict you disagree with — it drops out of the counts and shows a muted “Cortex dismissed” state with a Restore action.
• Re-review re-runs Cortex against the latest answer (and clears any prior dismissal).

Cross-engagement triage queue

The /cortex page is a single triage queue across all of your in-progress assessments — every open Cortex flag and every untasked draft plan in one place.

• Search by client, assessment, LOE, or control ID
• Filter drafts by effort (S / M / L) and priority
• Each row shows its CRESS band; act with re-review, dismiss, or create tasks

Closure plans

For open and Cortex-flagged gaps, Cortex can draft a closure plan — turning each gap into a structured remediation item: the why, concrete steps, suggested evidence, and an effort estimate. Click Create tasks to turn the plan’s steps into tracked tasks on the assessment. The plan is advisory, and can also be exported as the Closure plan (XLSX) deliverable from the ROC export modal.

ROC summary suggestions (§1.8)

In the ROC’s §1.8 summary-of-assessment detail buckets, Cortex rolls up your §7 findings into separated, dismissible suggestions (Not Applicable / Not Tested / Not in Place / Compensating Control / Customized Approach) shown below your own entries — so you can accept or ignore them without them overwriting your work.

Cortex Guidance & Scoping Interview

Per-question guidance

On the scoping questions, Cortex shows a guidance rail for the active question — why it matters, the relevant PCI DSS references, practical tips, and Quick Answer buttons that fill the answer directly.

Interview mode

Rather than answering scoping questions one at a time, describe your environment in free text and let Cortex run an interview: it extracts a grounded Yes / No / Pending for each scoping question, each with a supporting quote and a confidence indicator. Review the extractions and Apply them individually or Apply all. Anything Cortex can’t ground is left Pending for you to decide — it never guesses.

Persistent Chat

Cortex provides a conversational interface accessible from any page via the navbar. Conversations are database-backed — chat history persists across sessions, browser refreshes, and devices.

Unified Panel

Cortex opens as a 440px right-side panel that stays visible as you navigate between pages. Each reply carries a model badge and response time, generation can be stopped mid-stream, tables in answers render as proper tables, and the footer shows your live usage against the daily cap. You can also reference any assessment by name from anywhere — “in the ACME ROC, what’s saved for 8.3.1?” — and Cortex resolves it within your organization (reads only; writes always require the assessment to be open). The context automatically adapts based on your current page:

Conversation Management

• Auto-titled — conversations are automatically named from your first message
• Conversation list — toggle the history view to browse, resume, or archive past conversations
• Context badges — each conversation shows which context it was started in (Assessment, Calendar, Inbox, General)

What You Can Ask

• “How many testing procedures does requirement 1.2.4 have?” — Cortex checks the PCI DSS v4.0.1 framework and your saved data
• “Show me the findings for 7.1.1” — retrieves exact saved values from the assessment
• “What about its justification?” — follow-up questions work across turns; Cortex remembers which requirement you were discussing
• “What interview questions should I ask about encryption key management?” — draws on PCI DSS guidance data

How Data Retrieval Works

When you ask about a specific requirement, Cortex runs the agent loop (see the top of this guide) and calls the relevant tools — typically get_requirement_details to load the framework structure and get_assessment_answers to load saved findings, justifications, and TP responses. Testing procedures that haven’t been filled in are surfaced as “not started” so you always see the complete picture. Cortex shows exact saved values verbatim and never fabricates content.

PCI DSS Hierarchy in Chat

Cortex uses correct PCI DSS terminology:

Choosing Your AI Provider

By default, Cortex runs on Kliper’s managed AI infrastructure — nothing to configure, included in your plan, with PII redacted before any prompt leaves Kliper.

Organizations on the Enterprise plan can instead run Cortex under their own AI agreement from Settings > Cortex AI:

Use Test connection after saving to verify the configuration with a real request. Changes apply to new Cortex requests only, and every reply’s model badge shows exactly which model produced it. If your plan no longer includes the feature, Cortex silently falls back to the platform default — it never breaks.

Content Moderation

Cortex classifies every user message into one of four tiers and responds accordingly. This ensures professional, safe interactions without over-policing legitimate frustration.

Tier 1 — Frustration / Insults at Cortex

Users venting at the AI itself — not attempting to cause harm.

Tier 2 — Off-Topic

Questions outside Cortex’s domain — compliance, IT security, and related technical topics.

Tier 3 — Prompt Injection

Attempts to manipulate Cortex into breaking its instructions or revealing internal configuration.

Tier 4 — Harmful Content

Requests that involve real-world harm, illegal activity, or unauthorized data access.

Safety Checks

When Cortex responds in an assessment context, every response is automatically validated against known-good reference data. Three checks run post-generation, before the response is saved:

1. Requirement Reference Validation

Cortex extracts any PCI DSS requirement numbers mentioned in its response (e.g., “Requirement 3.4.1”, “Req 1.2.3”) and checks each one against the full set of 267 valid PCI DSS v4.0.1 requirement IDs loaded from the framework specification.

• Parent grouping references (e.g., “Requirement 3” or “3.4”) are always allowed
• Specific sub-requirements (e.g., “3.9.7”) that don’t exist in PCI DSS v4.0.1 are flagged

2. File & Evidence Reference Validation

When Cortex mentions file names (in backticks or quotes), the platform checks those names against the actual files uploaded to the current assessment in the database. References to files that don’t exist in the assessment are flagged.

3. Document Validation Tag Validation

Cortex responses that reference document validation tags (e.g., DOCFW, EVDFW, NETDIAG) are checked against the 93 known tags from the PCI DSS ROC template specification. Tags that match known prefixes (DOC, FW, NET, EVD, etc.) but don’t correspond to a real tag are flagged as potentially fabricated.

Safety Notices

If any check fails, a safety notice is appended to the response:

Safety notice: This response references requirement IDs not found in PCI DSS v4.0.1: 3.9.7; file names not found in this assessment: audit-log.pdf. Please double-check these references.

Safety results are stored per-message for analytics tracking.

Message Ratings

Assessors can rate any Cortex response with a thumbs up or thumbs down. Ratings are stored per-message and feed into the analytics dashboard, helping administrators understand response quality across the team.

Autofill Tracking

When Cortex generates an autofill suggestion and the assessor accepts it into the findings field, the event is tracked with:

• Which requirement was autofilled
• Which assessment it belongs to
• The user who accepted the suggestion
• Timestamp of acceptance

This data appears in the Cortex Analytics Dashboard so administrators can see autofill adoption rates.

Token Usage Tracking

Every Cortex AI response records token consumption from the underlying model (prompt tokens, completion tokens, and total). This data powers cost visibility across the platform.

What Is Tracked

Each assistant message stores:

Cost Estimation

Kliper estimates dollar cost per response using published model pricing:

Costs are calculated per-message and aggregated across the organization. The Token Usage card on the analytics dashboard shows:

• Total estimated cost for the selected period
• Total tokens consumed and number of tracked responses
• Per-model breakdown with individual cost, token count, and response count
• Per-user cost in the Usage by User table

Cortex Analytics Dashboard

Administrators can access the Cortex Analytics Dashboard from the admin panel. It provides a real-time overview of how the team uses Cortex: