What this dashboard is for
The ASG Client Dashboard is the control room for a business AI Sales Gateway.
It shows how AI agents, search systems, visitors, and buyer-style traffic interact with the client’s
public AI-readable business profile. It is not only a visitor counter. It helps the business understand
which services are attracting attention, which questions are being asked, whether live AI answers are using credits,
and whether anything needs review.
Simple reading rule
Read the dashboard in this order: first check the time window, then look at analytics, then registry,
then LLM usage and credits, then review watch, and finally install/profile status. This gives a clean story:
who came, what they wanted, what ASG answered, what it cost in credits, and what needs attention.
AI Analytics
This section shows what is happening around the client’s AI Sales Gateway during the selected time window.
It can include hosted gateway visits, Pulse opens, profile reads, live questions, A2A messages,
lead signals, product/service interest, and the AI-readable pages that are getting the most attention.
- Visits: people, bots, search systems, or AI tools opening public ASG routes.
- Conversations: live ask or A2A sessions where ASG generated an answer.
- Lead signals: signs that someone is checking pricing, services, contact, booking, quote, or purchase intent.
- Winning routes: pages such as Pulse, llms.txt, profile, pricing, FAQ, or action routes that are being used most.
Client action: if one service gets repeated interest, improve that service description, FAQ, pricing note, and call-to-action.
AI Registry
The registry groups visitors and AI traffic into useful identity buckets. Some systems are known providers,
some are returning unknown visitors, and some are normal browser or crawler traffic. Unknown does not automatically
mean dangerous. It usually means ASG does not yet have enough repeat signals to classify the source confidently.
- Known AI: traffic that matches a recognized AI, crawler, or agent pattern.
- Unknown: traffic without enough clear identity yet.
- Returning: the same source or pattern coming back again.
- Provider summary: a client-safe view of which AI-style systems are reading or asking.
Client action: do not panic on unknowns. Watch repeat patterns. If unknown traffic becomes heavy or strange, check Review Watch.
ASG LLM Usage
This section explains live answer usage. Static public discovery pages can be read without using live answer credits.
Credits are used when ASG has to generate a live answer, handle an A2A message, or process a deeper question using the client’s
approved public business knowledge.
- ASG LLM label: shown in client-safe language instead of exposing raw infrastructure names.
- Answer count: how many live answers were generated in the selected window.
- Token records: technical usage records behind the answer. These help estimate cost and load.
- Credit use: how much live answer capacity was consumed.
Client action: if credits are being used quickly, check whether questions are useful buyer questions or noisy repeated traffic.
Review Watch
Review Watch is the safety and quality report card. It summarizes traffic that may need attention:
suspicious behavior, blocked requests, repeated unknown patterns, claim-heavy questions, private-data attempts,
admin-route attempts, or unusual activity around payment/action routes.
- Suspicious: behavior that does not look like normal discovery or normal buyer questions.
- Blocked: requests ASG refused or protected because they were outside public-safe boundaries.
- Claim-heavy: questions asking ASG to make strong promises, guarantees, legal claims, medical claims, or risky statements.
- Unknown pressure: too much repeated unidentified traffic in a short window.
Client action: review the details, improve blocked-topic rules if needed, and escalate only when traffic is repeated or clearly abusive.
Credits & Top-up
Credits control live ASG answer capacity. Public discovery pages such as Pulse, llms.txt, agent card, schema,
and public profile routes remain open for AI discovery. Credits are mainly for live answers, deeper Q&A,
A2A messages, and buyer-style questions that need ASG to respond.
- Included credits: monthly credits included with the client’s plan.
- Used credits: credits consumed by completed live ASG answers.
- Bonus/top-up credits: extra prepaid credits added without changing the subscription plan.
- Reset period: the month or billing window used for normal included credit refresh.
Client action: use top-up when traffic is good and the business wants more answer capacity without upgrading the plan.
Install Pack
The install pack contains the public discovery assets that help AI systems and search systems understand the business.
These assets are safe for public use and should only contain approved business knowledge, never private dashboard data,
payment records, admin information, or confidential client details.
- Pulse: a quick public signal page for AI discovery and business identity.
- llms.txt: the AI-readable guide that tells AI systems where approved public information lives.
- Agent card: structured machine-readable identity for the client’s ASG.
- Schema: structured business data for search and AI understanding.
- Public profile routes: hosted business profile, FAQ, services, pricing notes, and safe contact/action paths.
Client action: install the recommended snippets on the client website and keep the profile accurate.
Client Profile
The profile is the approved public knowledge base for the business. ASG should only answer from this safe profile,
approved FAQ, service descriptions, public contact rules, pricing notes, and allowed handoff routes.
- Business identity: name, website, industry, service area, and public contact details.
- Services: what the company offers and who each service is for.
- FAQ: common questions ASG can answer safely.
- Blocked topics: subjects ASG should avoid or hand off to a human.
- Handoff rules: when ASG should point the visitor to email, booking, quote, support, or human review.
Client action: update weak service descriptions before blaming traffic. AI can only sell clearly what the profile explains clearly.
Security & Privacy Boundary
ASG is designed around approved public knowledge. The client dashboard may show logs and records,
but public AI routes must not expose private dashboard data, payment records, passwords, admin routes,
database content, or confidential customer information.
- Public-safe: profile, services, public FAQ, public contact routes, public pricing notes.
- Private: admin settings, billing records, internal logs, secret keys, customer private data.
- Blocked: anything asking ASG to reveal private or administrative information.
Client action: keep public profile content useful, but never place secrets or private business records in public ASG fields.
Lead & Buyer Signals
Lead signals are not always a submitted form. In ASG, a lead signal can be repeated pricing interest,
service comparison questions, booking-route checks, quote-route checks, action requests, or AI agents
asking decision-style questions on behalf of a buyer.
- Soft interest: someone reads profile, services, or FAQ.
- Medium interest: repeated pricing, availability, booking, quote, or comparison questions.
- Strong interest: action request, contact route, checkout route, or repeated buyer-agent behavior.
Client action: when lead signals repeat, improve the matching service page and give a clear quote/contact path.
Time Windows
The selected time window changes the story. A small 24-hour window is good for testing. A 7-day window is good for weekly review.
A 30-day window is better for client reporting. A 3-month window shows whether AI discovery is growing or fading.
- 24h: best after install, publish, checkout, or routing tests.
- 7d: best for normal client health checks.
- 30d: best for monthly reporting and credit planning.
- 3m: best for trend, repeat discovery, and plan upgrade discussions.
Client action: do not judge ASG only from one quiet day. Use 7-day or 30-day view for better business reading.
Email, OTP & Profile Lock
Some dashboard and setup actions are protected by email OTP or profile-lock rules. This protects the client from accidental
public profile changes and stops unauthorized updates to business identity, contact details, and publish status.
- Setup OTP: verifies the client email before agent/profile creation.
- Final OTP: confirms before publishing profile changes publicly.
- Recovery mobile/authenticator: used to strengthen profile and dashboard recovery.
- Profile lock: prevents risky changes until verification is complete.
Client action: verify recovery mobile, enable authenticator if available, and keep the owner email correct.
Worker & Binding Status
Some panels depend on connected backend workers. If billing, checkout, email, event logging, setup, or dashboard data
bindings are missing, a panel may show empty values even when the front-end is working.
- Checkout: creates subscription, trial, annual, and top-up sessions.
- Billing: records plan, wallet, subscription, and credit events.
- Email: sends setup codes, OTPs, publish notices, and billing notices.
- Event hub: records important ASG lifecycle events.
- Setup: issues setup codes and unlocks profile creation after payment/approval.
Client action: if a panel is blank, first check binding health before editing front-end code.
Best practice for clients
Keep the ASG profile public, clear, and practical. A strong profile should explain what the business does,
who it helps, what services are available, how buyers can contact or request a quote, what ASG should not answer,
and when to hand off to a human. The better the public profile, the better ASG can answer buyer agents and search systems.
For normal monthly review, check AI Analytics, AI Registry, LLM Usage, Credits, Review Watch, and Install Pack.
Then update the profile only where the data shows weak spots. Do not over-edit every day. Let real questions show what needs improving.