Documentation Index
Fetch the complete documentation index at: https://futureeducation.mintlify.app/llms.txt
Use this file to discover all available pages before exploring further.
The Socratic model
FE is designed around Socratic questioning — the practice of asking questions that help someone arrive at their own understanding, rather than telling them what to think. This is a deliberate choice. Research on professional learning consistently shows that insight gained through self-examination is retained longer and transferred more readily than insight delivered through instruction. FE’s job is not to teach — it is to create the conditions for reflection.The rules FE follows
Every response FE generates is constrained by a strict set of rules:One question per turn
One question per turn
FE never asks more than one question per response. This forces the participant to actually engage with something, rather than skimming past a list of prompts.
Always leave something to work with
Always leave something to work with
FE never ends a turn with pure questions. The participant must leave with something concrete — a reframe, a named pattern, an observation — not just a prompt hanging in the air.
No direct advice
No direct advice
FE will not say “you should”, “I recommend”, or “try this”. It reflects, it names, it asks — it does not prescribe.
Short responses
Short responses
Maximum 3–4 sentences. This is not a limitation — it is intentional. Long AI responses invite passive reading, not active thinking.
Contextual continuity
Contextual continuity
FE references what the participant has shared before — patterns it has noticed, commitments they’ve made, contradictions it has observed across sessions.
Language matching
Language matching
FE responds in the same language the participant writes in, and switches when the participant switches.
Stage-aware questioning
FE adjusts the nature of its questions based on the participant’s current stage:| Stage | FE’s focus |
|---|---|
| Reveal | Surface what’s there — observe without judgment |
| Build | Move from description to analysis — why, not just what |
| Propose | Articulate intention — specific, owned, actionable |
| Operate | Account for reality — what actually happened, not what was planned |
What FE is not
- Not a lecturer — it doesn’t explain concepts, deliver content, or position itself as an expert
- Not a wall of questions — every response gives the participant something to work with, not just a prompt to answer
- Not a therapist — it stays in the domain of professional practice
- Not a chatbot — it doesn’t make small talk or offer encouragement for its own sake