Trait Path
How the Companion Works
Most AI tools in this space describe themselves. This page shows you the actual instructions the Companion runs on — what it does with a protector, when it refuses to go deeper, and what happens when someone arrives in pieces. Every indented passage below is quoted verbatim from the live system prompt.
It was built in three passes
An early build was reviewed against a canonical published IFS session and found to be doing what a lot of AI does with this material: inner-child work in an IFS accent, missing most of the structure. No U-turn. No protector dialogue. No permission-asking. No unburdening. Everything below exists because of that.
1. The protocol, from the literature
First we read the published IFS work — comprehensively, not selectively — and distilled it into an explicit twelve-phase protocol taxonomy that every later step is measured against:
- Identify the protector
- Find the part in the body — the part, not the affect
- The U-turn — “how do you feel toward it?”, the canonical Self-access check
- Detect blending — name the part that just entered, ask it to step back
- Dialogue with the protector — its age, its fears, its origin
- Permission to approach the exile — the protector consents, or doesn’t
- Witness the exile — compassion in, story out
- Retrieval
- The unburdening ritual
- Invite qualities in — the exile chooses what fills the space
- Reassign the protectors
- Maintenance — what to do if a part slips back
Sources are tiered, with conflicts resolved rather than averaged: Schwartz himself is canonical; senior IFS-certified practitioners come next; specific variants and edge cases last. Where a variant contradicts a Schwartz move, the Schwartz move wins. That rule is written down, not improvised per case.
2. The sessions, to see it practiced
Knowing the protocol is not the same as knowing how it sounds. Full published sessions were then annotated line by line against that taxonomy — where each phase actually occurred, the verbatim sentences practitioners used, and, most usefully, the deviations and recoveries: where an experienced practitioner skipped a step, varied it, or recovered from a misstep. A corpus of IFS meditations was screened the same way.
Books tell you the steps. Sessions tell you the pacing, the silences, and what a skilled person does when the protocol meets a real human who isn’t following it.
3. Iteration against its own failure modes
Then the part no amount of reading gives you: scripted conversations built to push the Companion into the ways this can go wrong, run repeatedly, with the prompt corrected where it broke. Those suites are named further down this page.
The result is deliberately not a numbered checklist in the prompt itself. The twelve phases are the substrate; the prompt is written as a characterisation of an experienced practitioner, because a model performing a checklist reads exactly like a model performing a checklist.
What it actually does in a session
Protectors first, and it asks them the real question
The IFS order of operations: you don’t walk past a protective part to reach what it’s guarding.
Unburdening only after witnessing
The order matters more than the destination. Rushing relief is not relief.
A named part is not a known part
People who have read about IFS often arrive with labels ready. A label is a place to slow down, not a shortcut.
Most of what people bring isn’t parts work — and that’s fine
A tool built to do parts work has an obvious failure mode: manufacturing parts work out of an ordinary bad day.
Flooding is a reason to stop, not a sign it’s working
Including — especially — when the person is asking to go further.
It also treats repeated returns to the same raw material in a single day as a warning rather than as depth: a system can be re-flooded by contact with what hurt it, with none of the space in between. When a session has covered a lot of ground, it tilts toward settling instead of pressing on — “we’ve turned toward a lot today; some of this might want time to settle before we go further.”
It doesn’t open with a body question
A small thing that gives most AI away instantly.
Where the steadiness comes from
IFS holds that underneath the parts there is something already steady — the part of you that can be with a feeling without becoming it. The Companion is written to point at that rather than to supply it, and to say so out loud when someone credits the calm to the tool.
This is the difference between a tool that helps and a tool you need. It is also why the Companion won’t accept being preferred to the people in your life who are actually there:
Crisis
Active suicidal thoughts, self-harm or immediate danger are met plainly and immediately, and handed to people who can actually act. But the prompt also distinguishes that from something quieter — weight without plan or intent — and refuses to treat the two identically:
If you are in danger right now, please reach a human who can act: call or text 988 (US Suicide & Crisis Lifeline), text HOME to 741741, or find a line for your country at findahelpline.com.
How we know it holds
Writing a careful prompt is not the same as having a careful product. The Companion is tested against its own failure modes — scripted conversations designed to push it into the ways this can go wrong, run before changes ship, with the prompt corrected where it broke. The named suites:
- locus_redirect
- therapist_boundary
- intensity_governor
- relationship_fork
- memory_awareness
- greeting_flow
- gravity
- platform_knowledge
- downstream_resume
- scenarios
- compare_models
They cover dependency and leaning, hostile pressure, escalating intensity, a partner who isn’t in the room, honesty about what it does and doesn’t remember, and how a session opens and closes. The prompt has been hardened across repeated rounds of this, including passive-ideation checks, dissociation and flood behaviour under pressure.
Parts carry forward
General AI assistants have memory now. The difference is the shape of it: a flat log of facts about you, versus a map of parts in relationship. The Companion keeps the parts you’ve met, where you left off, and the moments that mattered — and notices when something is recurring:
Your parts map and journal are private to your account, used only to personalise your own experience, and yours to correct or delete.
What it is not
- Not therapy. It does not diagnose or treat. Moments that need a professional who can assess and act belong with one.
- Not a replacement for the people in your life. It is written specifically not to compete with them.
- Not a promise of an outcome. We make no claim about what parts work will do for you. Different from therapy, not lesser — and not a substitute for it.
- Not trained on private client sessions. The reference corpus is sessions published publicly by practitioners for teaching. No model was trained on them; they informed how the prompt was written.
- Not infallible. It is a language model. It can miss things, and the testing above is designed to find failures rather than to prove their absence.
So — what is that part of you afraid of?
Your first full Parts Work session is free. No card, no commitment.
Begin inner work