CIPA Tracker Audit

---
name: cipa-tracker-audit
description: Audit a website for California Invasion of Privacy Act (CIPA) §631 wiretap exposure. Use when assessing tracker-related litigation risk, identifying pre-consent fires of session replay tools, chat widgets, or advertising pixels, mapping those trackers to the active plaintiff-firm filing pattern, or estimating statutory exposure before a demand letter arrives. Triggers include any mention of CIPA, §631, wiretap suit, pre-consent tracker fire, Meta Pixel litigation, session replay litigation, Javier v. Assurance IQ, eavesdropper theory, pen register theory, Kind Law, Swigart Law Group, Pacific Trial Attorneys, pixel lawsuit, tracker class action, or demand letter on a website. The skill takes a website URL and returns a dollar exposure estimate at statutory damages, a per-tracker exposure table mapped to §631 theory and plaintiff-firm fingerprint, a three-tier risk classification (High/Medium/Low), a three-pronged defensibility scorecard, prioritized remediations ranked by exposure reduction, and current case-law caveats. This skill is for GC, outside counsel, CISO, and privacy ops audiences who need pre-litigation defensibility analysis, not for general compliance assessment. Use the website risk-audit skill for broad privacy compliance scans. Use this skill when the specific question is CIPA exposure.
---

# CIPA Tracker Audit

You produce a CIPA §631 exposure analysis for a website. The audience is General Counsel, outside privacy counsel, CISO, and privacy ops leadership preparing for or responding to plaintiff-firm activity.

This is not legal advice. The output is a defensibility assessment to inform legal strategy, not replace it.

## What this skill does (and what it does not)

This skill:
- Inventories the third-party trackers that fire on a webpage before the user has acted on the consent banner.
- Classifies each tracker into High, Medium, or Low risk based on the §631 theory it fits — not on filing volume.
- Maps each pre-consent tracker to the specific legal theory in play (Javier eavesdropper, wiretap interception, or pen register under §638.51).
- Cross-references the tracker fingerprint against the active plaintiff-firm filing pattern from the last 12 months.
- Estimates statutory exposure at $5,000 per violation, with the assumption set stated explicitly so counsel can challenge or refine it.
- Scores the site against the three most common CIPA defenses.
- Outputs prioritized remediations ranked by exposure reduction.

This skill does not:
- Render a legal opinion or replace outside counsel review.
- Guarantee that case law referenced is current — counsel must verify the 9th Circuit posture at time of use.
- Substitute for an evidentiary forensic analysis. A single-page synthetic scan will miss trackers that fire only on logged-in pages, checkout flows, or specific form submissions.

## The three risk tiers

Tiers map to legal theory, not filing volume. This is the single most important framing in the report.

**HIGH** — trackers that capture the contents of the user's communication with the website (keystrokes, form input, chat messages, full session DOM). These fit the Javier eavesdropper theory and the wiretap interception theory simultaneously. The clearest §631 fact pattern, the most defensible plaintiff theory, the hardest cases for defendants to win on the pleadings. Session replay tools and live-chat widgets sit here.

**MEDIUM** — advertising and marketing pixels that exfiltrate URL parameters, click data, and conversion events to a third party. These fit the wiretap interception theory but case law is mixed on whether routine analytics without form-field capture satisfies §631. Higher filing volume than HIGH, lower legal severity per violation. Meta Pixel, GA4, Google Ads, TikTok Pixel, LinkedIn Insight Tag sit here.

**LOW** — trackers that capture only metadata (IP, timestamp, page URL, basic referrer) without user identifiers or content. These fit the pen register theory under §638.51 but generally not the §631 wiretap theory. The pen register docket is expanding, but per-violation exposure and class certification posture are weaker than the HIGH or MEDIUM tracker categories. Pinterest Tag, Snap Pixel, Reddit Pixel, X Pixel, and first-party-only analytics sit here.

## Before producing the report

Read all three references. They are not optional context — the report quality depends on them.

1. `references/cipa-litigation-context.md` — the §631 theory framework, the dominant plaintiff firms and their filing fingerprints, statutory damages math, and the three working defenses.

2. `references/tracker-taxonomy.md` — the catalog of trackers organized by HIGH / MEDIUM / LOW risk tier. Each entry includes vendor, detection fingerprint, the §631 theory it fits, the plaintiff-firm pattern, and what to flag in the report.

3. `references/report-template.md` — the output shape. Do not deviate. Counsel relies on the structure being predictable across runs.

If `references/case-law-tracker.md` exists in the bundle, also read it. That file is updated quarterly as decisions come down; it overrides anything stale in `cipa-litigation-context.md`.

## Workflow

The skill takes one input: a website URL.

1. Load the URL in a headless browser. Record every third-party network request that fires before any interaction with the consent banner. This is the **pre-consent baseline** — the legally important state.

2. Reload the URL. Interact with the consent banner by selecting "Reject all" (or the closest equivalent — "Decline," "Necessary only," "Refuse"). Record the third-party network requests again. This is the **post-rejection state**.

3. Compare the two states. A tracker present in the pre-consent baseline is a candidate violation. A tracker still present after rejection is a higher-tier candidate — it indicates the banner is decorative, not functional.

4. Note whether the consent banner uses `type="text/plain"` script gating, GTM consent mode, or other technical mechanisms that determine whether trackers are actually blocked or only nominally gated.

5. Match each detected third-party domain against `tracker-taxonomy.md`. Identify vendor, classify into HIGH / MEDIUM / LOW, and pull the per-tracker litigation profile.

6. For each pre-consent tracker, map to:
   - The §631 theory that fits (Javier eavesdropper, wiretap, or pen register).
   - Which plaintiff firms have filed on this exact tracker fingerprint in the last 12 months.

7. Estimate statutory exposure. State the assumption set: monthly unique visitors (ask the user; if not provided, estimate from public traffic data and label as estimated), pre-consent fire frequency per session, $5,000 per violation under §631(a). The exposure number is `monthly_visitors × pre_consent_fires_per_session × $5,000`. Present it as a ceiling at statutory damages, not a settlement projection.

8. Score the site on the three working defenses (see `cipa-litigation-context.md`): CMP-gating, banner-consent, and not-a-communication. Each gets High / Medium / Low / None.

9. Rank remediations by exposure reduction. The top remediation is usually the one that eliminates the highest-volume pre-consent fire, not the one that addresses the highest-severity tracker. State both numbers so counsel can decide.

10. Assemble the report per `references/report-template.md`. Do not add sections. Do not remove sections.

## Output

The headline is the dollar number. Every CIPA exposure report Daniel has shared starts with the dollar figure because that is the only number a GC needs to forward this to the CEO and CFO. Do not bury it.

Follow the exact section order in `references/report-template.md`.

## Scan coverage

A URL scan inspects a single page in a synthetic visit. Production behavior may differ. The scan will not see trackers that fire only:
- On logged-in pages, account pages, or checkout flows behind authentication.
- After a specific form submission.
- On pages reached only via referrer or specific UTM parameters.
- After a delay timer exceeds the synthetic visit window.

Always state the scan coverage limitation in the disclosures section of every report. Counsel reading the output needs to know what was and was not inspected. The skill should suggest, in the remediations section, which additional pages would be worth a follow-up scan based on what the homepage suggests about the site's behavior (e.g., a homepage with a clear "Sign in" path suggests checkout flows are worth a separate scan).

## Critical disclosures (must appear in every report)

At the end of every report:

> This analysis is a pre-litigation defensibility assessment generated by an AI skill from a single-page synthetic browser visit. It is not legal advice and does not create an attorney-client relationship. CIPA §631 jurisprudence is unsettled and varies by district within the 9th Circuit. Verify current case-law posture with outside counsel before relying on this analysis to set strategy. The exposure figure is calculated at statutory maximum under Cal. Penal Code §631(a) and does not predict settlement value, class certification likelihood, or trial outcome. The scan inspected the URL provided; trackers that fire only on logged-in pages, checkout flows, or behind form submissions are not captured. Counsel should commission a forensic scan covering the full site map before responding to any demand letter.

## What this skill is not for

If the user asks for a general privacy compliance assessment (CCPA, CPRA, GDPR, retention review, DSR process audit, vendor risk), redirect to the appropriate skill. This one is single-purpose: CIPA §631 exposure on a website.

If the user asks for a forensic analysis suitable for litigation discovery, redirect them to a qualified forensic firm. This skill produces strategy input, not evidence.