Upload today
Upload PDF and Markdown today. More formats are coming.
Corpus
It starts with everything your organization already knows. It stays true.
01
Upload, clipper, Confluence, Teammate, and any MCP agent. Same checks. Same permissions.
Upload PDF and Markdown today. More formats are coming.
The Chrome clipper saves only the page you choose. Confluence imports as a copy with a link back.
Tell an agent to add this to Cluesora. Ask Teammate in the app, Slack, or Teams.
Your wiki and search stay. Cluesora sits beside them.
Same checks for every way in
Dead ends and agent updates are written back
02
Decisions and their reasons, sources, people, systems, what to learn first, and data tables.
What you decided, and why, stays findable by intent, not only by the words on a page.
Teams, services, vendors, customers, and who can actually answer, not only who is mentioned.
For every idea it knows what must be understood first.
Tables stay as rows you can query and total. Pages cite them live.
03
A question is matched to what your organization actually decided and meant.
It also matches keywords and meaning, but intent is the difference.
“Can a new enterprise customer go monthly?” finds “Why we bill annually.” The question never says annually.
Enterprise pays yearly. Customers on monthly plans left at renewal. Billing cannot prorate mid-term.
Asked“Can a new enterprise customer go monthly?”
Why we bill annually.
Enterprise customers pay yearly.
04
Every result shows the exact line, its source, why it matched, and what it connects to.
You see the line that answers you, not a file to go hunt through.
Each result says why it matched: by intent, by meaning, or by the words.
Related ideas, owners, and other sources sit next to the line.
Asked“Who signs off when a new laptop costs more than usual?”
Hardware above the standard budget needs the department head’s approval.
05
When the story is spread across several places, Cluesora follows connections it has already mapped.
The reading is done before anyone asks, so a question lands on the answer in one hop.
It does not guess the path from page to page while you wait.
An edge case in a second document arrives with the main answer, cited.
06
When sources disagree, you see both sides and where each came from.
One page says refunds take 14 days, three others say 30. Both stay visible.
Nothing is blended into a confident wrong answer.
When sources agree, you get one answer with all of them cited. A caveat is kept.
A tool picks 30 days because more pages say it, and hides the 14-day page.
You see 14 days and 30 days, each with its source.
14 days
30 days
07
Routine upkeep is automatic. Unclear calls get checked first. Anyone can correct what is wrong, and their edit wins.
At a dead end it finds the person who knows, asks them, and writes their answer back.
Agents update the existing doc instead of filing a second version.
New versions take over. Old ones are kept and dated, never deleted.
Small updates are usable within about a minute. Large PDFs take up to ten minutes.
Enterprise P1s: first response within one hour.
Written back from Slack by the support leadNo answer on record.
The question hit a dead end08
When something your content needs is not taught anywhere yet, Cluesora keeps a list.
It shows exactly what your library assumes but never explains.
Fifty topics assume basic Python and nothing teaches it. Cluesora lists it once.
When a Python primer is added, all fifty connect to it.
09
Bad input cannot bend the map. After a big change it checks its own work and fails loudly.
Garbled scans are held back until fixed. A passage repeated fifty times counts once.
Generic headings do not glue unrelated things together.
Every page gets a short summary and a longer study summary. Both rewrite when content changes.
It checks its own work after every big change and fails loudly instead of quietly getting worse.
10
The same map answers a person in the app and an agent in a coding tool.
A person gets the answer, the exact lines, the sources, and the conflicts.
Agents search first, then save what they settle. Reads are never metered.
Nobody, person or agent, sees more than the source allows.
No. When content arrives it is read, summarised, connected, and checked. Nobody tags or files anything.
The question is matched to what you decided and meant, not only to the words on a page.
You see both sides and where each came from. Nothing is blended into one wrong answer.
Yes. Tables stay as rows you can query and total. Pages cite them live.
No. Agents update the existing doc. Sources are never rewritten.