How to Generate a Briefing

Generating a briefing is straightforward: ask a question, read the sourced response. This page walks through the full process and explains what you'll see in the output.

Step by Step

Navigate to the BRIEFINGS tab in the top navigation bar. You need to be signed in to access it.

Type your research question into the input box. Be as specific as you like — the more focused the question, the more precise the answer.

Press Enter or click the send button. The AI begins searching the document library and composing a response.

The response streams in as it is generated. You can start reading immediately — you don't need to wait for the full answer to finish.

Review the completed briefing, including the inline citations and the References section at the bottom.

Inline Citations: Blue and Orange Chips

Inside every briefing response you will see two types of highlighted chips embedded in the text:

Blue chips represent entities — people, organizations, programs, or locations that are named in the library. Clicking a blue chip opens the full entity profile in a detail panel on the right side of the screen.
Orange chips represent events — specific incidents or documented occurrences. Clicking an orange chip opens the event detail, including date, description, and the source documents behind it.

These chips are how the AI shows its work. Every claim is grounded in something specific, and the chips let you follow the chain of evidence directly from the briefing text to the underlying source.

The References Section

At the bottom of every briefing is a References section listing all the source documents the AI drew on to construct the response. Each reference links to the original document in the library. Use this section to verify the sources, read the full context, or assess the evidence tier of what you just read.

Token Cost

A token estimate is shown before you submit a question, and the actual cost is displayed after the briefing completes. The cost depends on the complexity of the question and how many library documents the AI needs to synthesize. Your remaining balance is always visible in the account area. See Tokens & Credits Explained for full details.

Tips for Good Questions

Be specific about your subject. "What did Luis Elizondo say about AATIP before Congress?" will return a tighter, better-sourced answer than "tell me about AATIP."
Ask research questions, not opinion questions. The AI works from document evidence. Questions like "what is documented about X" play to its strengths better than "do you think X is real."
Name the people and programs you care about. Proper nouns give the AI clear targets to search for in the library.
Ask one focused thing at a time. Compound questions ("tell me about Roswell and the Tic Tac and AARO") produce less focused responses than separate targeted questions.

Tip: After reading a briefing, use the Follow-Up / Branch button to dig deeper into something it mentioned. Branching is cheaper than a fresh question because the context is already loaded. See Branching: Follow-Up Research.

When the AI Says It Doesn't Have Enough Information

If the AI responds that there isn't enough in the library to fully answer your question, try a few things:

Rephrase the question using different terminology — for example, try an acronym instead of the full name, or vice versa.
Check the SOURCES page to see whether the topic you're researching is covered in the library at all.
Use SEARCH to find related documents directly, then run a more targeted briefing question based on what you find.

The library is built from primary sources and does not include everything ever written about UAP — if a source hasn't been ingested, the AI won't have it.