Continue the AIBound landing-pages + tracking workstream (Vercel ad-landers side). The 2 landers are BUILT, refined, and LIVE on fgabranden Vercel; only tracking + cleanup remain. Work through the remaining tasks below. ⚠️ A PARALLEL SESSION IS LIVE — DO NOT CLOBBER IT. A separate session is rebuilding the MAIN aibound.com site in Webflow (Lovable → Webflow). Hard boundaries: - Touch ONLY /Users/flori/fga-site-aibound-landers + the tracking vault docs (landing-pages-and- webflow-app-buildplan.md, tracking-audit.md, linkedin-qualified-lead-capi-buildplan.md). - NEVER touch /Users/flori/fga-aibound-webflow, the lovable-to-webflow-* vault docs, the Webflow Designer, or apex aibound.com. Domain split: apex = Webflow (theirs), go.aibound.com = Vercel (mine). - BROWSER: use chrome-devtools MCP (its OWN dedicated profile, already logged into Google/GTM/LinkedIn as branden@flyinggoatagency.com) OR Playwright headless. NEVER use claude-in-chrome, NEVER select_browser Browser 2 (deviceId b4fa7565-200b-45b9-9cef-0c859ef42517 = the Webflow session's Chrome), and NEVER pkill/kill any Chrome / chrome-devtools / claude-in-chrome process. If a browser tool won't attach, use Playwright headless or wait — do not kill shared processes. - GTM container GTM-P7KCBTS6 is SHARED (that session is NOT editing it): keep GTM edits ADDITIVE (new trigger + new tags only, never edit/delete existing tags) and PING ME before Publishing. - Shared vault docs: re-read immediately before every write, targeted edits only, expect modified-since-read collisions. READ FIRST: obsidian-fga/system/clients/aibound/landing-pages-and-webflow-app-buildplan.md §10–12 (esp. §11 NEXT-SESSION HANDOFF + the ANTI-CLOBBER note), tracking-audit.md, and linkedin-qualified-lead-capi-buildplan.md. STATE: landers live at https://fga-site-aibound-landers-pi.vercel.app/{prevent-high-risk-ai, find-your-gaps} (fgabranden Vercel, auto-deploy on git push; repo pushes as fgabranden via the repo-scoped cred store — do not change git identity). go.aibound.com CNAME is WITH THE CLIENT at Dynadot — do NOT block on DNS; SITE_URL auto-swaps when it lands. REMAINING TASKS, in order: 1. GTM booking-conversion tracking (do first; not DNS-blocked). The landers push dataLayer event `schedule_meeting` on a completed HubSpot booking (src/components/MeetingScheduler.tsx). In GTM container GTM-P7KCBTS6 add a Custom-Event trigger on `schedule_meeting` → fire Meta Pixel `Schedule` (dataset 3784273851714143), LinkedIn "Booked Demo" (Insight tag 9647708), Google Ads booking conversion (AW-18091099342). Stage in a workspace, verify via GTM Preview + a REAL test booking (each fires once; Meta Events Manager shows Schedule), show me the results, then ask before Publish. 2. Priority-1 unified n8n CAPI workflow (the original critical path). Follow n8n ops rules (reference_n8n_stack_guide, reference_n8n_patch_protocol, reference_n8n_http_pagination, reference_n8n_api_put_settings_allowlist) + the §10 patch protocol in obsidian-fga/system/CLAUDE.md. Build DEACTIVATED: HubSpot contact → lifecyclestage `marketingqualifiedlead` (Ops-Hub webhook if scope allows, else Schedule + HubSpot Search-API cursor on the MQL-transition date) → Code node → fan out to (a) LinkedIn CAPI POST https://api.linkedin.com/rest/conversionEvents (conv URN urn:lla:llaPartnerConversion:27253436, SHA256 lowercased email, conversionHappenedAt = MQL epoch ms, $0; headers LinkedIn-Version + X-Restli-Protocol-Version:2.0.0 + Bearer; mint token at build via Campaign Manager acct 531000256 → /buyer-actions-manager/sources/SERVER_API → Generate access token) and (b) Meta CAPI POST /v21.0/3784273851714143/events with META_TOKEN_AIBOUND. continueOnFail:true; fire ONLY on the transition INTO MQL. Creds HUBSPOT_TOKEN_AIBOUND + META_TOKEN_AIBOUND are in fga-ai-demo/.env.local (HubSpot token VERIFIED, polling path unblocked; app client-secret only needed for signed webhooks). Test the data shape → SHOW ME the exact payloads → my approval → activate. 3. Task 8 gcloud cleanup: I'll run `! gcloud auth login` in the prompt first. Then delete SA aibound-tracking-audit@floribot-491223.iam.gserviceaccount.com + its token-creator binding, revoke ADC, brew uninstall gcloud, verify the prod host is clean. Record in tracking-audit.md #8. 4. When the client confirms the Dynadot CNAME is in: dig go.aibound.com to verify → Vercel auto-issues SSL → set up GA4 cross-domain (aibound.com ↔ go.aibound.com) → confirm Meta domain aggregation covers the subdomain → hand the branded go.aibound.com URL to Spencer. Write back to the vault continuously (targeted edits). Ask before any GTM Publish, n8n activate, or other outward action. That's the single prompt — all remaining landers/tracking work, fully fenced off from the Webflow session.
AIBound

Your Control Plane for Secure AI

Browser · Endpoint · Network · Cloud

AIBound is the AI Security Control Plane purpose-built for enterprise security teams. Discover and eliminate shadow AI, build your complete AI inventory, and enforce AI governance policy — from endpoint to cloud.

24 HoursRead‑OnlyNo New Agents
ChatGPTGitHub CopilotCursorGeminiClaudeMidjourneyDALL·ECopilotGrammarlyNotion AIJasper AIHubSpot AICanva AIAdobe FireflyDeepSeekCopy.aiPerplexityAnthropicOpenAITensorFlowChatGPTGitHub CopilotCursorGeminiClaudeMidjourneyDALL·ECopilotGrammarlyNotion AIJasper AIHubSpot AICanva AIAdobe FireflyDeepSeekCopy.aiPerplexityAnthropicOpenAITensorFlow
AWS BedrockAzure OpenAIGoogle Vertex AIDatabricksLM StudioCrewAICodeiumv0.devmonday AIGongPower BI CopilotJetBrains AIBraveChorus.aiWeights & BiasesDataikuAgentGPTIntercom FinSlack AISalesforce EinsteinAWS BedrockAzure OpenAIGoogle Vertex AIDatabricksLM StudioCrewAICodeiumv0.devmonday AIGongPower BI CopilotJetBrains AIBraveChorus.aiWeights & BiasesDataikuAgentGPTIntercom FinSlack AISalesforce Einstein
Finding Shadow AI is enough.

Shadow AI is just the start.The real risk is what it does next.

AIBound doesn't just find AI. It traces what each one can do, scores the risk, and stops the dangerous ones. Automatically.

1

Discover AI Footprint

Inventory
2

Expose AI Identities

Identity + skills
3

Map Data Connections

Exposure
4

Measure AI Risk

Risk A–F
5

Enforce in Real Time

Real-time

90 seconds · How AIBound stops high-risk AI in 5 steps

Book a Demo

We Catch Shadow AI Risks Before Impact.So You Can Innovate Faster.

Empower your security team to stay ahead of new AI risk so you can push AI adoption forward with confidence.

01

One Click to Connect

Agentless. Read-only. No software to install.

Connect in one click, get insights in hours. 100+ integrations across code, cloud, endpoints, networks, and security tools.

One Click to Connect
01

One Click to Connect

Agentless. Read-only. No software to install.

Connect in one click, get insights in hours. 100+ integrations across code, cloud, endpoints, networks, and security tools.

One Click to Connect
02

See It All

Your complete AI inventory — visible in hours.

Track AI apps, agents, plugins, models, and MCPs the moment they appear. Visibility across endpoints, networks, cloud and code.

See It All
02

See It All

Your complete AI inventory — visible in hours.

Track AI apps, agents, plugins, models, and MCPs the moment they appear. Visibility across endpoints, networks, cloud and code.

See It All
03

Know What It Does

Understand what each AI tool touches.

Reveal true intent, permissions, and data access patterns. Know what every agent can reach and whether it's operating within policy.

Know What It Does
03

Know What It Does

Understand what each AI tool touches.

Reveal true intent, permissions, and data access patterns. Know what every agent can reach and whether it's operating within policy.

Know What It Does
04

Know If It's Safe

Trust scores your organization can act on.

Assess the security posture, data handling, and business risk of every AI resource. Know which to trust and which to restrict.

Know If It's Safe
04

Know If It's Safe

Trust scores your organization can act on.

Assess the security posture, data handling, and business risk of every AI resource. Know which to trust and which to restrict.

Know If It's Safe
05

Prevent It — In Real Time

Set guardrails with Guardian AI and enforce them instantly.

Block shadow AI, coach users, or prevent unauthorized tools on the spot. No tickets, no delays. Trigger immediate alerts.

Prevent It — In Real Time
05

Prevent It — In Real Time

Set guardrails with Guardian AI and enforce them instantly.

Block shadow AI, coach users, or prevent unauthorized tools on the spot. No tickets, no delays. Trigger immediate alerts.

Prevent It — In Real Time
The Guardian Engine

The Intelligence Layer Behind Every Insight

The Guardian engine fuses telemetry from your existing Browser, Endpoint, Network, and Cloud tools into meaningful AI risk signals, continuously translating noisy activity into ranked, actionable insights for security teams.

AI Applications
ChatGPT
OpenAI
Claude
Gemini
Copilot
Make
Replit
Perplexity
Cursor
n8n
Cohere
DeepSeek
Copy.ai
Databricks
Dataiku
MLflow
GitHub Copilot
Asana
Azure OpenAI
Intercom Fin
Brave
monday AI
Grafana
Tableau
LM Studio
Telemetry Sources
Browser
Chrome
Okta
Cloudflare
Endpoint
Palo Alto
Snyk
Jira
Network
Cloudflare
Palo Alto
GCP
Cloud
GCP
GitLab
GitHub
Intelligence Layer

Guardian

Engine

Insights Engine

100,000+ AI apps cataloged

Scoring Engine

Instant A–F risk grades

Swarm Engine

Proactive threat prevention

Outcomes
Fix the Right Risks First
See Every AI Resource, Everywhere
Stop Threats Before Damage
Board-Ready in One Click

Your security stack wasn't built for AI. You're not alone.Security leaders are feeling the blind spots every day.

"What sold me was connecting AIBound in an afternoon and seeing real AI risk the same day. No agents, no outages—just instant signal."

— Global CISO, Healthcare Enterprise

"For the first time, I can answer a simple question: where is AI actually running? Across vendors, side projects, and extensions, AIBound turned rumor into a living inventory."

— VP Information Security, B2B SaaS

"We discovered more AI usage in 48 hours than in the previous year of interviews, audits, and Slack archaeology. AIBound surfaced the shadow estate we all suspected."

— Chief Information Security Officer, Retail & E‑Commerce

"It's not just a list of tools; AIBound shows what each AI system actually touches—data, identities, workflows—so I can explain risk in business language, not stack diagrams."

— Director of Information Security, Financial Services

"The A–F risk grades changed the conversation with my board. Instead of 'AI is scary,' I show exactly which models are safe to scale and which need containment."

— Chief Privacy & Security Officer, Consumer Brand

"Our biggest gap was separating experimentation from exposure. AIBound's scoring made it obvious which AI pilots were harmless and which were quietly processing regulated data in production."

— Head of Security & Trust, AI SaaS

"Before AIBound, AI incidents meant days of forensics just to find the owner. Now high‑risk assets come with context and a playbook, so my team can move in minutes."

— VP Security Operations, Global Manufacturing

"Shadow AI used to feel like whack‑a‑mole. AIBound finally gave us a radar: unapproved tools, data paths, and users in one place—with the ability to coach, contain, or kill."

— CISO, Regulated Healthcare Technology Provider

Works With the Stack You Already Own

Read-only. No new agents. One-click connect. AIBound layers on top of your existing endpoint, network, and cloud tools.

JAMF
CrowdStrike
SentinelOne
Intune
Cloudflare
Palo Alto
Zscaler
Okta
Azure AD
GitHub
GitLab
Snyk
AWS
GCP
Azure
Splunk
Jira
Slack
ServiceNow
Chrome Enterprise
JAMF
CrowdStrike
SentinelOne
Intune
Cloudflare
Palo Alto
Zscaler
Okta
Azure AD
GitHub
GitLab
Snyk
AWS
GCP
Azure
Splunk
Jira
Slack
ServiceNow
Chrome Enterprise

Endpoint

Discover every AI app, extension, and agent installed across managed devices.

Network

Detect AI service traffic at the DNS and gateway layer — even when nothing is installed.

Code

Visibility into AI-powered IDE extensions, copilots, and code generation tools.

Cloud

Monitor AI workloads, models, and services across your cloud environments.

Identity

Map AI tool usage to users, roles, and departments for policy enforcement.

Browser

Track AI tools accessed through enterprise and personal browsers.

Workflow

Push findings into your existing ticketing, SIEM, and orchestration tools.

Governance

Feed AI risk data into your GRC and compliance platforms automatically.

How AIBound Goes Live in One Day

AIBound ingests your telemetry, builds your AI inventory, and starts enforcing policy. All in one working day.

01

Connect to Your Stack

Secure OAuth or API connection to your existing security and IT tools. No agents, no root access, no network taps.

02

Auto-Discover AI

Guardian ingests telemetry and auto-builds your AI inventory across browsers, endpoints, network, and cloud. First results show up in minutes.

03

Turn Signal into Enforcement

Approve, coach, or block AI tools directly through your existing controls. One click in AIBound, enforcement everywhere.

See Every AI Tool Across Your Fleet. Today.

Join leading teams to full AI visibility in under 24 hours.

Get a complete AI inventory in 24 hours — no agents, no network taps, no disruption.

SOC 2 Type II CertifiedISO 27001 Certified