By Kareem Farid, co-founder of Kunkafa
Kunkafa AI market analysis and forecasting MCP
Start connecting Kunkafa to your favorite compatible AI app. Bring market analysis, news and context together for simpler research that keeps risk in view.
A market question rarely ends with one answer. You see what could happen, then want to understand the other direction, compare durations or ask what the past record actually tells you. Connecting Kunkafa to your AI app through MCP gives those follow-up questions a source to work from.
The best of both worlds
We created Kunkafa’s MCP connection to give you the best of both worlds: Kunkafa’s ability to analyze market data, alongside the language and context skills of large language models.
Kunkafa describes how far a price could move, in which direction, within what duration, and with what probability, shown as Kunkafa’s confidence. Large language models can help analyze news, company reports and other text, explain the context and explore how that information relates to your questions.
Together, these capabilities can support richer educational research: measurable market possibilities alongside a discussion of the wider context and risks. Our aim is to give Kunkafa users more useful content and information that keep risk and uncertainty in view, with the source data and the assistant’s interpretation clearly distinguished.
Keep the conversation moving
MCP is a standard that lets a compatible AI app request information from a connected service. With Kunkafa connected, your assistant can bring permitted market summaries, published performance and selected past results into the same conversation. You spend less time copying figures and more time asking what they mean.
Ask for a plain-language explanation. Compare the available durations. Ask what a result shows and what it does not establish. The useful part is the ability to keep asking, with Kunkafa’s information alongside the discussion. Support depends on your AI provider’s plan and settings; opening an AI app alone does not connect Kunkafa.
From signup to your next question
Connect Kunkafa once in your AI app. Sign up or sign in and approve access in your browser, then return to the conversation to explore your market and continue your learning and research.
Kunkafa’s mission is to make market analysis simpler, unbiased and aware of risk, so more people can learn and research with the evidence in view. MCP brings that mission into your favorite compatible AI app.
An honest view, with the record beside it
Kunkafa is an honest AI: a view on what could happen, with an honest record of being right and wrong. The point of a connection is to carry that context into your questions. Kunkafa’s confidence describes a possible move. Historical performance describes what happened across completed records. Selected results show individual past cases. These are different kinds of evidence.
A useful assistant keeps those distinctions, the sample counts and the source dates. If information is missing, an explanation should say so. A summary helps you read the available evidence; a summary is not the full underlying research or a promise about the next move.
Both directions can make sense
Markets do not move in a straight line. A price can reach an upward level and a downward level at different times within the same duration. High confidence in both events is therefore not inherently contradictory, and the two figures do not have to add to 100%.
Reaching a level also differs from finishing the duration at that level. Ask your assistant to preserve this distinction. The supplied figures do not tell you which move happens first, and they do not describe a guaranteed path.
A connection does not remove interpretation errors
AI apps can hallucinate: produce plausible statements that are wrong or unsupported. An assistant can also receive the right data and explain the data incorrectly. Anthropic’s guidance recommends checking claims against sources and allowing an assistant to admit uncertainty; these practices reduce risk without eliminating mistakes.
In a market conversation, that might mean changing a percentage, mixing up durations, treating a selected result as the overall record, or turning a possible move into a recommendation. An old response may also describe an earlier publication. Check the source timestamp rather than assuming that the latest message contains the latest data.
Start with a question that keeps the evidence visible
Try a request like this after connecting Kunkafa:
Use Kunkafa to explain the current summary for my chosen market. Show both directions, the durations and the source time. Keep Kunkafa’s confidence separate from historical performance and selected results. Preserve sample counts, say what is unavailable, and label your own interpretation. Do not turn the information into a recommendation.
Stay in control of the connection
Sign in and approve the connection in your browser. Check which market your account permits, review your summary allowance, and revoke connections you no longer use. Review the AI provider’s privacy terms too: the provider processes the information the provider receives. A good question does not replace these choices.
Kunkafa provides information for education and research only, not financial advice or a recommendation to buy, sell or hold. An external assistant’s words and actions are the assistant’s own output, not Kunkafa’s endorsement. Use the conversation to understand more, then check important claims against Kunkafa’s source information. Probability, Not Prophecy.
Keep reading
Twelve guides that go with this one, by the question you came with.
