AI Agent (S.A.R.G.E.) reference

The AI Agent is a standalone AI workspace in UserSearch that can drive OSINT searches conversationally rather than card-by-card, and can also run a full-page entity search builder. It is branded S.A.R.G.E. — Strategic Autonomous Research Gathering Engine — and its in-chat persona is “Sarge, your senior investigator.”

The AI Agent is a documented product feature (per Lee, 2026-07-08), even though its interface carries a Version: 0.3 (experimental) label.

Note

AI Agent is both a View mode of the Dashboard and the feature described on this page. This reference covers the feature. Where the distinction matters, this page writes “the AI Agent view mode” versus “the AI Agent.”

The feature spans four surfaces:

Surface	What it is
Entity search builder	Full-page module multi-select, Case attach, and run
Chat landing	First entry point: composer, model selector, quick-start buttons
Consent modal	First-run privacy and responsible-use gate with two acknowledgements
Active chat	A live conversation thread

---

Entity search builder

The AI Agent exposes a full-page search builder that lets you compose a structured search before, or instead of, chatting.

AI Agent, entity search builder, Email entity active

Entity-type and Search type selection

The builder opens with two rows of tabs.

1. Row 1 selects the entity type: Username, Name, Email, Phone, Image, Domain, IP, VIN, Address, and Company. In the captured frame the Email entity is active, highlighted in the coral accent with a check-mark.
2. Row 2 selects the Search type: Corporate, Court, Cyber, Leaks, Messaging, and Wireless. The Cyber Search type carries a “+12” badge.

Caution

Some entity tabs carry a numeric “+N” badge (Domain shows “+2”, IP shows “+1”). These are inferred to indicate additional entity types grouped under that tab, but that behaviour is not yet confirmed.

The search-builder card

The builder card for the active Email entity contains:

• A header EMAIL with a mail icon.
• A query input with placeholder text i.e. [email protected].
• A collapsible ADDITIONAL PARAMETERS disclosure beneath the input, revealing further query-refinement fields.
• An Email OneScan summary line reading 5 modules selected.
• A CASE field with a Select a case dropdown, used to attach the search to an existing Case.
• A Search for Free button beside the Case selector.

Note

Why the button reads “Search for Free.” A OneScan search costs the sum of all the data sources left selected in the data-source selection, so the price moves as data sources are toggled on or off; a zero-cost composition therefore runs free (per Lee, 2026-07-08). By contrast, Search type searches other than OneScan are fixed price.

Module grid — Email modules

The builder shows the same Module catalogue you would see in the normal Email Search type. The Email OneScan tile is the selected Module, shown with a coral border and a row of data source icons. OneScan is the recommended multi-source Module that fans out across several data sources and merges the results.

The individual Email Modules and their data sources, as read from the tiles:

Module	data source	What it does
Email Search	UserSearch (usersearch.ai)	Check what social networks and dating sites an email is registered on.
Email Search	OSINT.INDUSTRIES (osint.industries)	Find enriched social media networks linked to the email address.
Email Search	PredictaSearch (predictasearch.com)	Find enriched social media networks linked to the email address.
Email Search	EPIEOS (epieos.com)	Find enriched social media networks linked to the email address.
ProtonMail Analyse	ProtonMail (proton.me)	Confirm if an email uses proton mail, locate the creation date and sometimes its forwarding email.
Email-To-Name	Gravatar (gravatar.com)	Determine a real name and location of a person, by email.
Scam Database	Global Scam Database (scamsearch.io)	Check if an email has been reported in a romance, crypto, or general scam.
Domain Ownership	UserSearch (usersearch.ai)	Determine what website domains are owned by the email address.
People Search	PIPL (pipl.com)	Find enriched business and social data on a person from a combination of details.
Public Leaks	IntelX (intelx.io)	Search for publicly leaked information relevant to your query.
Public Leaks	Dehashed (dehashed.com)	Search for publicly leaked information relevant to your query.
Public Leaks	HavelBeenPwned (haveibeenpwned.com)	Review reported password breaches on HavelBeenPwned, for a particular email address.
Email Threat	HudsonRock (hudsonrock.com)	Identify computers that are comprimised by hackers or virus which are used by the email address.

Note

The data source name “HavelBeenPwned” and the word “comprimised” are reproduced exactly as they appear in the product. Do not correct their spelling when quoting the control.

Each Module tile carries an info (i) icon in the top-right and a small circular status icon in the bottom-right. Selecting a tile’s (i) opens details about that Module and data source.

---

Chat landing

The AI Agent’s chat entry point presents S.A.R.G.E. as an “AI Investigation Agent.”

AI Agent, chat landing, composer and quick-start buttons

Identity and status

• Page title: AI Agent.
• A circular monitor-glyph avatar with an orange gradient sits above the heading.
• Heading: AI Investigation Agent.
• Subtitle: Powered by S.A.R.G.E. (Strategic Autonomous Research Gathering Engine).
• Two status pills sit side by side: a green Agent Online pill and a Version: 0.3 (experimental) pill.

The composer

• Placeholder text: I’m Sarge, your senior investigator. Give me a lead and I’ll run it down.
• An Image attachment control at the bottom-left.
• A model selector at the bottom-right reading Claude Opus 4.8 with a chevron. Opening it reveals alternative model choices.
• A send button at the far right.

Caution

SargeBot is model-agnostic (per Lee, 2026-07-08): the selector lets you choose the AI provider (Claude, GPT, Grok, and others) and version, and swap it at will. Claude Opus 4.8 is the model selected in the captured screens.

Quick-start buttons

Two rows of one-click starters seed the composer or launch a search type:

• Entity starters: Email, Username, Domain, Phone Number, Search the Web.
• Action starters: Ask a Question, Give an Opinion, Full Search.

A Dashboard → link sits below the starters.

Caution

The exact resulting state of each quick-start button, and the destination of the Dashboard → link, have not been captured.

Chat history sidebar

The left sidebar contains a CHAT HISTORY section with a help ? icon. On the landing frame it lists a LAST 7 DAYS group and an OLDER group, plus a View all history link showing a count. The global sidebar also carries New Chat and Cases nav items; when empty, the body reads No chat history yet.

---

First-run consent modal

Before you can use S.A.R.G.E., a first-run modal titled Before You Use S.A.R.G.E. (shield icon, sub-label PRIVACY, DATA & RESPONSIBLE USE NOTICE) gates access. It presents five notice sections and requires two acknowledgements.

AI Agent, "Before You Use S.A.R.G.E." consent modal

Notice sections

1. YOUR CONVERSATIONS STAY ON YOUR DEVICE — Conversations are not stored on UserSearch servers. All chat history is saved locally on your computer and encrypted. You can delete your conversation history at any time by clearing your browser cache, or directly on this page.
2. THIRD-PARTY AI SERVICES — The assistant is powered by third-party AI data sources including Grok, GPT, and others. Your messages and data are transmitted to these services to generate responses. Do not use the service if you do not want your conversations accessible to third parties.
3. AI-POWERED RESPONSES — All responses from S.A.R.G.E. are generated by third-party artificial intelligence and may contain inaccuracies, outdated information, hallucinations, or errors. You take full responsibility for this.
4. VERIFY ALL INFORMATION — Search results and agent output should be independently verified. Do not rely solely on AI-generated content for critical decisions.
5. RESPONSIBLE USE — The tool is intended for lawful OSINT research only. You are responsible for ensuring your use complies with applicable laws and ethical guidelines.

Caution

The consent modal’s “Grok, GPT, and others” and the composer’s Claude Opus 4.8 are consistent, not contradictory: SargeBot lets you choose the AI provider and version (per Lee, 2026-07-08). The modal lists selectable AI providers; the composer shows the current selection.

Note

The on this page deletion link performs an in-app deletion of local conversation history. Its exact location and behaviour have not been captured.

Required acknowledgements

A REQUIRED ACKNOWLEDGEMENTS panel contains two unchecked checkboxes:

• Acknowledgement 1: “I understand that AI-generated responses may be inaccurate and I will verify information independently before acting on it.”
• Acknowledgement 2: “I confirm I am using this service for lawful purposes only. I accept full responsibility for how I use the information provided and will not hold UserSearch liable for any outcomes resulting from my use of this service.”

A progress counter reads 0 OF 2 CONFIRMED. The footer has two buttons, Decline and Accept & Continue.

Caution

Until both boxes are checked, Accept & Continue is inferred to be disabled, and the result of selecting Decline has not been captured. Neither behaviour is yet confirmed.

---

Active chat

Once past the consent gate, an active thread renders as a conventional chat.

AI Agent, active S.A.R.G.E. chat thread

• Page title: AI Agent.
• A right-aligned user message bubble and an assistant reply.
• The bottom composer is focused with a coral outline, and carries the same placeholder, Image attachment control, Claude Opus 4.8 model selector, and send button as the landing.
• Below the composer: a caution line AI-generated responses may be inaccurate and a Feedback link.
• The CHAT HISTORY sidebar adds a TODAY group with the active thread highlighted, and the View all history count rises, consistent with a new conversation having been created.

---

Credits — the separate S.A.R.G.E. pool

UserSearch maintains two separate credit wallets (per Lee, 2026-07-08): the main Global search-credits pool and the Sarge (AI Agent) pool. They are topped up and spent independently.

The AI Agent surfaces this split directly:

• Two Credit pills sit top-right: Global and Sarge, each with its own balance.
• A note under the composer states: Chat & reports use Sarge credits · searches use Global credits, followed by a How credits work link.

So within the AI Agent, chat turns and generated reports draw from the Sarge pool, while any underlying searches the agent runs draw from the Global pool.

Caution

Across the captured surfaces, the Global pill reads $0.00 on the entity search-builder frame but $22.88 on the chat frames. Whether this reflects the two-pool model, a different session or unfunded state, or a stale render has not been resolved. Do not treat either explanation as fact. The destination of the How credits work link has also not been captured.

Verified against UserSearch v2.0.20