Prompt details

Tested on

Claude Opus 5.5

Length

525 words · 4 fill-ins

Example input

[De-identified session jottings]: Sess 7. 50 min, individual, in person. Adult client, panic disorder + GAD (dx per intake). CBT. - GAD-7 today 14 (sess 1: 18, sess 4: 16) - Panic attacks: 2 since last sess, both on AM commuter train, peak 7/10 and 6/10. Was ~4/wk at intake. - Breathing log 4 of 7 days. Skipped weekend, "felt fine so didn't bother" - Avoidance: took bus instead of train 2 days. Still avoiding express train. - Reviewed log. Cog restructuring on "I'll faint and no one will help": evidence for/against. C came up with "I've had 30+ attacks and never fainted." - Interoceptive exposure: spinning in chair 30 sec x3, SUDS 70 > 55 > 40. Hyperventilation 60 sec x2, SUDS 65 > 50. - C tense at start, fidgeting, more relaxed by end, laughed once. - Work stress up, new manager. Said "sometimes I wonder what the point of any of it is" re job. Didn't follow up, ran out of time. - Sleep 5-6 hrs, says improving. - HW: breathing daily; ride local train 2 stops x3 this week with SUDS ratings; thought record x2. - Next: review exposures, hierarchy step 3 (express train 1 stop). Consider sleep hygiene.

[Note format and template]: SOAP. Our agency template: S, O, A, P headings; first line gives session number, length and modality; A must state progress toward treatment plan goals; P must include the next appointment. Under 300 words.

[Your discipline and setting]: Licensed professional counselor (LPC), outpatient group practice, adult clients, sessions billed to commercial insurance.

[Treatment plan goals]: Goal 1: Reduce panic attacks from about 4 a week to 1 or fewer a week within 12 weeks (client's panic log). Goal 2: Reduce GAD-7 from 18 to under 10 within 16 weeks. Goal 3: Ride the commuter train to work 4 of 5 days a week without getting off early (exposure hierarchy).

Example output

Session 7 | 50 min | Individual, in person | CBT

Client reported 2 panic attacks since last session, both on the morning commuter train, peak 7/10 and 6/10. Client reported breathing practice 4 of 7 days and skipped the weekend: "felt fine so didn't bother." Client reported taking the bus instead of the train on 2 days and continues to avoid the express train. Client reported higher work stress with a new manager and said, about the job, "sometimes I wonder what the point of any of it is." Client reported sleeping 5–6 hrs per night and described sleep as improving.

GAD-7: 14 (session 1: 18; session 4: 16). Client appeared tense and fidgeted at start, appeared more relaxed by end, and laughed once. Clinician reviewed the log. Cognitive restructuring on "I'll faint and no one will help" (evidence for/against). Client generated: "I've had 30+ attacks and never fainted." Interoceptive exposure: chair spinning 30 sec ×3, SUDS 70 → 55 → 40; hyperventilation 60 sec ×2, SUDS 65 → 50. Client's statement about "the point of any of it" was not explored because time ran out.

Panic disorder; GAD (per intake).

Goal 1 (panic ≤1/wk): 2 attacks since last session vs ~4/wk at intake. Progressing, not met.

Goal 2 (GAD-7 <10): 18 → 16 → 14. Progressing, not met.

Goal 3 (train 4/5 days): panic continues on the commute, bus substituted 2 days, express avoided. [clinician to complete: train days ridden this week; current hierarchy step]

SUDS decreased across interoceptive trials. Client independently generated disconfirming evidence. Breathing practice inconsistent when symptoms are low. [clinician to complete: clinical impression, including work stress and the "point" statement]

Homework: daily breathing practice; ride local train 2 stops ×3 with SUDS ratings (Goal 3); thought records ×2 (Goals 1–2).

Next session: review exposures; hierarchy step 3 (express train, 1 stop); consider sleep hygiene.

Follow-up: [clinician to complete: plan to explore the "point of any of it" statement / risk assessment]

Next appointment: [clinician to complete: date/time]

---

Flags for the clinician

Risk: "sometimes I wonder what the point of any of it is" was not followed up. The jottings have no risk screen. Assess and document before signing, or note when you will assess.
Risk: the jottings record no suicide/self-harm screen this session. The note does not state that risk was denied.

Claude Opus 5.5

SOAP Progress Notes – SOAP note prompt for ChatGPT and Claude

Q: Can ChatGPT write SOAP notes?

Yes, ChatGPT can turn session jottings into a well-structured SOAP note. Use de-identified jottings only, unless your organisation has approved the tool for client information, and review every line, since you sign the note and the model knows only what you wrote.

Q: Is it HIPAA compliant to use ChatGPT for progress notes?

Pasting identifiable client information into a consumer chat app is generally not HIPAA compliant unless the provider has signed a business associate agreement with your practice. De-identified jottings lower the risk; check your organisation's policy and use an approved tool for anything identifiable.

Q: What is the difference between SOAP, DAP and BIRP notes?

SOAP splits a note into Subjective, Objective, Assessment and Plan. DAP folds the first two into Data, then Assessment and Plan. BIRP, common in behavioral health, records Behavior, Intervention, Response and Plan. The prompt writes whichever format your agency uses.

Q: Can AI write the risk assessment in a therapy note?

No, and this prompt does not try. A risk assessment rests on your clinical judgment and on what you asked in session, so the prompt never writes that a client denied risk unless your jottings say so, and flags every risk item for you to assess and document.

^$9.99

Kept in your account
Copy filled in or download
Updates included

Pay once and keep it. Or get every premium prompt and skill with the plan, $19.99 a month. Payments are non-refundable. See our Terms.

votes views

Claude Opus 5.5

This prompt turns your de-identified session jottings into a SOAP, DAP or BIRP progress note. Paste your jottings, template, setting and goals; you get the note, flags for risk items and placeholders, and what your jottings missed.

What you get

A progress note in SOAP, DAP or BIRP, following your template
Client reports, your observations and your interpretation kept apart, tied to treatment goals
Flags for every risk item and placeholder you still need to complete
What a reviewer or payer may look for that your jottings did not record

How to use it

1
Remove every identifier
Strip names, dates of birth, places and anything else that could identify the client, and check your organisation's AI policy first.
2
Paste jottings, template and goals
Add your template rules and the treatment plan goals so the note shows progress in your own measures.
3
Complete the flags yourself
Assess and document every risk flag and placeholder, edit the note into your voice and sign it in your records system.

Better results

Paste one of your own past notes, de-identified, with the jottings and ask it to match that style.
Jot your risk screen during the session. The prompt will not write a risk assessment you did not record, so it can only flag the gap.
Follow up with "Rewrite this as DAP" if your agency or payer uses a different format.

Questions

Can ChatGPT write SOAP notes?

Yes, ChatGPT can turn session jottings into a well-structured SOAP note. Use de-identified jottings only, unless your organisation has approved the tool for client information, and review every line, since you sign the note and the model knows only what you wrote.

Is it HIPAA compliant to use ChatGPT for progress notes?

Pasting identifiable client information into a consumer chat app is generally not HIPAA compliant unless the provider has signed a business associate agreement with your practice. De-identified jottings lower the risk; check your organisation's policy and use an approved tool for anything identifiable.

What is the difference between SOAP, DAP and BIRP notes?

SOAP splits a note into Subjective, Objective, Assessment and Plan. DAP folds the first two into Data, then Assessment and Plan. BIRP, common in behavioral health, records Behavior, Intervention, Response and Plan. The prompt writes whichever format your agency uses.

Can AI write the risk assessment in a therapy note?

No, and this prompt does not try. A risk assessment rests on your clinical judgment and on what you asked in session, so the prompt never writes that a client denied risk unless your jottings say so, and flags every risk item for you to assess and document.

Updated 8 Nov 2026