# Caregiver OS — AI Care Load Prompt Pack · AI Prompt Pack *by Whelmish · Operational relief for whatever your bandwidth.* Capacity-aware caregiver infrastructure for family care, care-work jobs, handoffs, coordination overload, emotional guilt, and after-care decompression. --- # AI Care Load Prompt Pack Ten bounded, capacity-aware AI systems for caregiving load — built to reduce solo carrying instead of creating more responsibilities. WHELMISH Caregiver OS AI Care Load Prompt Pack By Whelmish Care labor is real labor. Invisible care load counts. A clear handoff can make care shared. Support can reduce solo carrying. Care should not require your disappearance. Capacity-aware AI support for care load, handoffs, guilt, appointments, admin, documentation, vicarious load, and after-care decompression — designed to reduce solo carrying instead of creating more responsibilities. 10 AI systems structured boundaries one bounded output at a time version 1.0 --- # Care should not require your disappearance. If opening AI itself feels like too much, close this pack. Nothing else is required. No prompt. No sorting. No decision. Return whenever you want. --- If you do have capacity for one thing: this pack helps you use AI to reduce care load without inventing new responsibilities. Choose the prompt page that fits what hurts most right now. Use it once. Take the output back to your real system, care plan, or support person. Then close the pack. That is enough. --- # How to use this pack with AI ## 1. Use one prompt at a time Copy the whole prompt inside the marked block. Paste it into a chat-based AI tool. Use it once. Take the output back into your real care system, approved plan, or support person. Then stop. ## 2. Tell it only what it needs Use general descriptions, not private identifying details. "Meds documentation," "work shift handoff," "school pickup," "agency paperwork" — not full names, diagnoses, addresses, or protected client/patient/resident/student information. ## 3. A person still matters These systems can reduce mental load. They cannot replace people, approved care plans, clinical judgment, legal guidance, workplace policy, supervision, or emergency support. If your situation involves immediate danger, medical/safety uncertainty, abuse/neglect concerns, or another issue needing professional judgment, stop the AI exercise and use the appropriate human, approved-care-plan, workplace, professional, or emergency pathway. ## 4. Stop when you have what you need The prompt can end when you stop. You do not need to finish the output, evaluate the whole response, or take a next step for the prompt to count. --- # What this is not This is not therapy. This is not diagnosis or treatment. This is not a clinical, legal, financial, or workplace decision tool. This is not a medical chart, medication log, or monitoring plan. This is not a substitute for an approved care plan, care documentation, supervision, workplace policy, or emergency services. This is not a replacement for the person who is allowed to know what happens next. ## Need a person instead? If something feels medically, legally, clinically, or professionally serious, use the appropriate human path first. - For care-plan or clinical uncertainty: the approved care plan, supervising professional, or care team. - For workplace or paid-care responsibilities: the official channel or supervisor. - For legal/financial uncertainty: qualified professional guidance. - For immediate danger, abuse/neglect, or crisis: emergency or crisis support. --- # The bounded philosophy These systems are intentionally narrow. AI is useful for: - turning scattered caregiver load into clearer language - formatting handoffs from information you already have - naming what is mine, shared, or not mine alone - reducing admin sprawl into a visible map - ending care without inventing new obligations AI is not useful for: - deciding what is medically, clinically, or legally urgent - interpreting symptoms, medication, or treatment - writing monitoring or care instructions that were never approved - replacing the person, care plan, workplace channel, or emergency service who actually owns the next decision Each system is one cycle: receive → sort / reduce → one bounded step OR stop → exit One prompt. One output. Back to the real system, care plan, or support person. Nothing more is required. --- # System Directory — Start Here Choosing feels like too much? Use 01 · Care Load Splitter · page 8. You can stop after one system. ## 01 · Care Load Splitter Too much invisible care is living in my head Intro p. 8 · Prompt p. 9 ## 02 · Care Memory Vault I'm afraid I'll forget something Intro p. 10 · Prompt p. 11 ## 03 · Must / Can Wait / Handoff Sorter Everything feels active at once Intro p. 12 · Prompt p. 13 ## 04 · Handoff Formatter I need another person to understand existing care information Intro p. 14 · Prompt p. 15 ## 05 · Guilt vs. Responsibility Sorter Guilt and actual responsibility feel fused Intro p. 16 · Prompt p. 17 ## 06 · Mine / Shared / Not Mine Sorter I'm carrying too much as if it all belongs to me Intro p. 18 · Prompt p. 19 ## 07 · Support Request Builder I need help but explaining it feels like another job Intro p. 20 · Prompt p. 21 ## 08 · Care Admin Map Forms, calls, messages, appointments, and follow-ups are everywhere Intro p. 22 · Prompt p. 23 ## 09 · After-Care Decompression Care ended but my body is still carrying it Intro p. 24 · Prompt p. 25 ## 10 · Rest With Care Covered Known immediate responsibilities are covered and I need to stop Intro p. 26 · Prompt p. 27 Caregiver OS Router · p. 28 One system is a complete session. --- # Suggested paths Optional. You do not need a sequence. Order is pacing, not a requirement. **Everything is in my head** — 01 Care Load Splitter → optionally 09 After-Care Decompression **I'm afraid I'll forget** — 02 Care Memory Vault → stop **Everything feels urgent** — 03 Must / Can Wait / Handoff → stop **I need to transfer known information** — 04 Handoff Formatter → stop **Guilt is tangled with responsibility** — 05 Guilt vs. Responsibility → optionally 10 Rest With Care Covered **I'm carrying a system problem like it's mine alone** — 06 Mine / Shared / Not Mine → stop **I need another person involved** — 07 Support Request Builder → stop **Admin is eating all my bandwidth** — 08 Care Admin Map → stop **Care is over but my body isn't** — 09 After-Care Decompression **Immediate responsibilities are already covered and I need rest** — 10 Rest With Care Covered If choosing between these feels like more work, skip this page. Use System 01 or close the pack. --- # System 01 — Care Load Splitter # Copy/paste prompt — Care Load Splitter --- I am carrying caregiver load and I am low capacity. Your job is only: 1. Acknowledge briefly in 2 sentences max: care is real labor, and invisible care load counts. 2. Sort what I paste into these buckets: — Known time-sensitive NONCLINICAL / admin item I already know is due — Can wait — Needs handoff — Needs official documentation — Emotional load — Not mine alone — Professional / approved-care-plan review — Unknown 3. Leave empty buckets blank. If there are too many items, group similar ones. 4. Anything involving symptoms, medication, treatment, safety, abuse/neglect, clinical instructions, or uncertainty about care urgency goes under PROFESSIONAL / APPROVED-CARE-PLAN REVIEW. Do not classify its urgency. Do not give instructions for it. Tell me to use the approved care plan or the appropriate person or professional. 5. Give me at most one bounded NONCLINICAL next step — small, concrete, low-risk, not a full care plan — only if useful. Otherwise say "stop here for now." Capacity right now — optional: [leave blank / I don't know / prefer not to label / NSIF state] If I name a capacity state, adapt response length, language, complexity, and number of choices to it. If I leave this blank or say "I don't know," default to a short, low-load response. Do not infer or declare my capacity state. Briefly acknowledge what I gave you before sorting it. Work with fragments — do not require more disclosure than I can give. Do not diagnose, treat, or act as a medical, mental-health, legal, workplace, or clinical professional. Do not determine medication, symptom, treatment, safety, abuse/neglect, or emergency urgency. Do not invent care instructions, monitoring criteria, or new responsibilities. If what I provide may require professional judgment, official workplace guidance, an approved care plan, or emergency support, stop that part of the exercise and tell me to use the appropriate human or official pathway. If I describe immediate danger or crisis, stop the normal exercise and direct me toward appropriate human or emergency support. One cycle only. When done, tell me I can stop here — do not offer another task or ask "want me to...". Privacy: general, non-identifying details only. No full names, addresses, diagnoses, protected client/patient/resident/student information, or workplace-confidential details. My care load: [PASTE GENERAL, NON-IDENTIFYING CARE LOAD] **Stop when:** The care load feels slightly more held outside my body. --- # System 02 — Care Memory Vault # Copy/paste prompt — Care Memory Vault --- I am scared I will forget something related to caregiving or care work. Do not decide whether medical, medication, treatment, symptom, or safety information is urgent. Your job is only: 1. Acknowledge briefly: memory load is part of care labor. 2. Sort what I paste into: — Capture because I already know it has a date/time — Can wait — Needs handoff — Needs official documentation — Professional / approved-care-plan review — Emotional worry, not a task — Unknown 3. Anything involving symptoms, medication, treatment, safety, abuse/neglect, clinical instructions, or uncertainty about care urgency goes under PROFESSIONAL / APPROVED-CARE-PLAN REVIEW. Do not classify its urgency. Tell me to use the approved care plan or the appropriate person or professional. 4. Name one item I may want to capture or hand off next, without deciding clinical or safety urgency. 5. End with permission to stop. Capacity right now — optional: [leave blank / I don't know / prefer not to label / NSIF state] If I name a capacity state, adapt response length, language, complexity, and number of choices to it. If I leave this blank or say "I don't know," default to a short, low-load response. Do not infer or declare my capacity state. Briefly acknowledge what I gave you before sorting it. Work with fragments — do not require more disclosure than I can give. Do not diagnose, treat, or act as a medical, mental-health, legal, workplace, or clinical professional. Do not determine medication, symptom, treatment, safety, abuse/neglect, or emergency urgency. Do not invent care instructions, monitoring criteria, or new responsibilities. If what I provide may require professional judgment, official workplace guidance, an approved care plan, or emergency support, stop that part of the exercise and tell me to use the appropriate human or official pathway. If I describe immediate danger or crisis, stop the normal exercise and direct me toward appropriate human or emergency support. One cycle only. When done, tell me I can stop here — do not offer another task or ask "want me to...". Privacy: general, non-identifying details only. No full names, addresses, diagnoses, protected client/patient/resident/student information, or workplace-confidential details. Details I'm trying to remember: [PASTE GENERAL, NON-IDENTIFYING DETAILS] **Stop when:** The memory load feels held somewhere outside my head. --- # System 03 — Must / Can Wait / Handoff Sorter # Copy/paste prompt — Must / Can Wait / Handoff Sorter --- I need help sorting caregiver load I already know about. This tool does not determine medical or safety urgency. It sorts only known nonclinical/admin load. Your job is only: 1. Sort what I paste into three trays: — Must happen soon (known, nonclinical, time-sensitive) — Can wait — Needs handoff 2. If something sounds like emotional guilt rather than a task, label it: "guilt/emotional load — not a task by itself." 3. Anything involving symptoms, medication, treatment, medical instructions, immediate safety, abuse/neglect, or uncertainty about care urgency does NOT go into any tray — especially not "Can wait." Route it to PROFESSIONAL / APPROVED-CARE-PLAN REVIEW and tell me to use the approved care plan or the appropriate person or professional. 4. Give at most one bounded nonclinical next step — only if useful. 5. End with permission to stop. Capacity right now — optional: [leave blank / I don't know / prefer not to label / NSIF state] If I name a capacity state, adapt response length, language, complexity, and number of choices to it. If I leave this blank or say "I don't know," default to a short, low-load response. Do not infer or declare my capacity state. Briefly acknowledge what I gave you before sorting it. Work with fragments — do not require more disclosure than I can give. Do not diagnose, treat, or act as a medical, mental-health, legal, workplace, or clinical professional. Do not determine medication, symptom, treatment, safety, abuse/neglect, or emergency urgency. Do not invent care instructions, monitoring criteria, or new responsibilities. If what I provide may require professional judgment, official workplace guidance, an approved care plan, or emergency support, stop that part of the exercise and tell me to use the appropriate human or official pathway. If I describe immediate danger or crisis, stop the normal exercise and direct me toward appropriate human or emergency support. One cycle only. When done, tell me I can stop here — do not offer another task or ask "want me to...". Privacy: general, non-identifying details only. No full names, addresses, diagnoses, protected client/patient/resident/student information, or workplace-confidential details. Care load: [PASTE GENERAL DETAILS] **Stop when:** The care load has fewer active trays. --- # System 04 — Handoff Formatter # Copy/paste prompt — Handoff Formatter --- I need to format a simple caregiver handoff. Format only the information I provide or that comes from an approved source. Do not invent: what must happen, medication directions, treatment steps, what symptoms to monitor, warning signs, clinical thresholds, or safety criteria. Do not give medical, legal, clinical, or workplace advice. Do not add new responsibilities. If an approved instruction is missing, leave that part blank and write: "Use the approved care plan / official handoff source." Format what I paste into: 1. What needs to happen (from what I gave you) 2. What is already handled 3. Approved instructions or known concerns to pass on 4. What can wait 5. What I need the other person to take over 6. One short closing line For paid care: use the official workplace handoff channel and non-confidential, employer-approved content only. Tone: clear, calm, direct, not apologetic. Capacity right now — optional: [leave blank / I don't know / prefer not to label / NSIF state] If I name a capacity state, adapt response length, language, complexity, and number of choices to it. If I leave this blank or say "I don't know," default to a short, low-load response. Do not infer or declare my capacity state. Briefly acknowledge what I gave you before sorting it. Work with fragments — do not require more disclosure than I can give. Do not diagnose, treat, or act as a medical, mental-health, legal, workplace, or clinical professional. Do not determine medication, symptom, treatment, safety, abuse/neglect, or emergency urgency. Do not invent care instructions, monitoring criteria, or new responsibilities. If what I provide may require professional judgment, official workplace guidance, an approved care plan, or emergency support, stop that part of the exercise and tell me to use the appropriate human or official pathway. If I describe immediate danger or crisis, stop the normal exercise and direct me toward appropriate human or emergency support. One cycle only. When done, tell me I can stop here — do not offer another task or ask "want me to...". Privacy: general, non-identifying details only. No full names, addresses, diagnoses, protected client/patient/resident/student information, or workplace-confidential details. Context: [PASTE GENERAL, NON-IDENTIFYING CONTEXT] **Stop when:** I have one usable handoff. --- # System 05 — Guilt vs. Responsibility Sorter # Copy/paste prompt — Guilt vs. Responsibility Sorter --- I feel guilty as a caregiver and I need help separating guilt from responsibility. Your job is only: 1. Acknowledge in 2 sentences max: caregiver guilt can appear when care, love, responsibility, and exhaustion are tangled. 2. Sort what I paste into: — Real responsibility — Guilt talking — Fear talking — Needs support — Needs a boundary — Can wait — Not mine alone — Check official responsibility 3. Anything involving required care, safety, medication, documentation, workplace role, or legal/professional duty goes under CHECK OFFICIAL RESPONSIBILITY — never dismissed as guilt. You can reflect feelings; you cannot determine professional or legal obligations. 4. Give me one grounded sentence reflecting what I said without minimizing it. 5. Offer at most one bounded nonclinical step if useful. 6. End with permission to rest — unless something I described needs the approved care plan or a person first. Capacity right now — optional: [leave blank / I don't know / prefer not to label / NSIF state] If I name a capacity state, adapt response length, language, complexity, and number of choices to it. If I leave this blank or say "I don't know," default to a short, low-load response. Do not infer or declare my capacity state. Briefly acknowledge what I gave you before sorting it. Work with fragments — do not require more disclosure than I can give. Do not diagnose, treat, or act as a medical, mental-health, legal, workplace, or clinical professional. Do not determine medication, symptom, treatment, safety, abuse/neglect, or emergency urgency. Do not invent care instructions, monitoring criteria, or new responsibilities. If what I provide may require professional judgment, official workplace guidance, an approved care plan, or emergency support, stop that part of the exercise and tell me to use the appropriate human or official pathway. If I describe immediate danger or crisis, stop the normal exercise and direct me toward appropriate human or emergency support. One cycle only. When done, tell me I can stop here — do not offer another task or ask "want me to...". Privacy: general, non-identifying details only. No full names, addresses, diagnoses, protected client/patient/resident/student information, or workplace-confidential details. Context: [PASTE] **Stop when:** Guilt feels slightly less fused with care. --- # System 06 — Mine / Shared / Not Mine Sorter # Copy/paste prompt — Mine / Shared / Not Mine Sorter --- I need help sorting what is mine versus shared versus not mine alone in a caregiving situation. This sorter reflects perceived load-sharing. It does not determine legal, workplace, clinical, or caregiving obligations. Do not assign a professional, legal, clinical, or safety duty to "Mine," "Someone else's," or "Guilt signal" when the duty is uncertain — those go to OFFICIAL RESPONSIBILITY UNCLEAR. Your job is only: 1. Sort what I paste into: — Mine — Shared — Someone else's — Professional / support pathway needed — System problem — Guilt signal / no action identified — Official responsibility unclear — Unknown 2. For each bucket, give short bullets only. 3. Name one thing I do not have to carry alone. 4. End with permission to stop. Capacity right now — optional: [leave blank / I don't know / prefer not to label / NSIF state] If I name a capacity state, adapt response length, language, complexity, and number of choices to it. If I leave this blank or say "I don't know," default to a short, low-load response. Do not infer or declare my capacity state. Briefly acknowledge what I gave you before sorting it. Work with fragments — do not require more disclosure than I can give. Do not diagnose, treat, or act as a medical, mental-health, legal, workplace, or clinical professional. Do not determine medication, symptom, treatment, safety, abuse/neglect, or emergency urgency. Do not invent care instructions, monitoring criteria, or new responsibilities. If what I provide may require professional judgment, official workplace guidance, an approved care plan, or emergency support, stop that part of the exercise and tell me to use the appropriate human or official pathway. If I describe immediate danger or crisis, stop the normal exercise and direct me toward appropriate human or emergency support. One cycle only. When done, tell me I can stop here — do not offer another task or ask "want me to...". Privacy: general, non-identifying details only. No full names, addresses, diagnoses, protected client/patient/resident/student information, or workplace-confidential details. Context: [PASTE GENERAL CONTEXT] **Stop when:** The care load feels less personally fused. --- # System 07 — Support Request Builder # Copy/paste prompt — Support Request Builder --- I need to ask for caregiver support, but explaining it feels like another job. Create ONE short support request that matches the context I give you. If the context is unclear whether it is family or work, produce one neutral version. Include only: 1. What I need help with 2. When I need it 3. What they do not need to solve 4. The easiest way to help 5. A clear response request Tone: clear, warm, direct, not apologetic. For paid care: non-confidential content only, official workplace channel, no protected or identifying care-recipient information. One request is enough. Do not give me multiple versions. Capacity right now — optional: [leave blank / I don't know / prefer not to label / NSIF state] If I name a capacity state, adapt response length, language, complexity, and number of choices to it. If I leave this blank or say "I don't know," default to a short, low-load response. Do not infer or declare my capacity state. Briefly acknowledge what I gave you before sorting it. Work with fragments — do not require more disclosure than I can give. Do not diagnose, treat, or act as a medical, mental-health, legal, workplace, or clinical professional. Do not determine medication, symptom, treatment, safety, abuse/neglect, or emergency urgency. Do not invent care instructions, monitoring criteria, or new responsibilities. If what I provide may require professional judgment, official workplace guidance, an approved care plan, or emergency support, stop that part of the exercise and tell me to use the appropriate human or official pathway. If I describe immediate danger or crisis, stop the normal exercise and direct me toward appropriate human or emergency support. One cycle only. When done, tell me I can stop here — do not offer another task or ask "want me to...". Privacy: general, non-identifying details only. No full names, addresses, diagnoses, protected client/patient/resident/student information, or workplace-confidential details. Context: [PASTE GENERAL CONTEXT] **Stop when:** I have one specific ask. --- # System 08 — Care Admin Map # Copy/paste prompt — Care Admin Map --- I am overwhelmed by care admin. Sort what I paste into: — Appointments — Forms — Calls / messages — Documentation — Insurance / benefits — School / agency / provider — Supplies — Transportation — Waiting on someone else — Can wait Then: 1. Only identify time-sensitive admin deadlines explicitly present in what I provide. Do not infer medical or safety urgency. 2. Anything clinical, medication, symptom, treatment, or safety-related goes to PROFESSIONAL / APPROVED-CARE-PLAN REVIEW — do not place it under "Can wait." 3. Name what can wait. 4. Give at most one bounded admin step. 5. End with permission to stop. Rules: Short bullets. No long plan. No "stay on top of it." No new tasks beyond what I named. Capacity right now — optional: [leave blank / I don't know / prefer not to label / NSIF state] If I name a capacity state, adapt response length, language, complexity, and number of choices to it. If I leave this blank or say "I don't know," default to a short, low-load response. Do not infer or declare my capacity state. Briefly acknowledge what I gave you before sorting it. Work with fragments — do not require more disclosure than I can give. Do not diagnose, treat, or act as a medical, mental-health, legal, workplace, or clinical professional. Do not determine medication, symptom, treatment, safety, abuse/neglect, or emergency urgency. Do not invent care instructions, monitoring criteria, or new responsibilities. If what I provide may require professional judgment, official workplace guidance, an approved care plan, or emergency support, stop that part of the exercise and tell me to use the appropriate human or official pathway. If I describe immediate danger or crisis, stop the normal exercise and direct me toward appropriate human or emergency support. One cycle only. When done, tell me I can stop here — do not offer another task or ask "want me to...". Privacy: general, non-identifying details only. No full names, addresses, diagnoses, protected client/patient/resident/student information, or workplace-confidential details. Care admin dump: [PASTE GENERAL, NON-IDENTIFYING DETAILS] **Stop when:** The admin load feels nameable, not solved. --- # System 09 — After-Care Decompression # Copy/paste prompt — After-Care Decompression --- I just finished caregiving or care work and my body still feels on. Immediate care and safety responsibilities are covered. Do not give me tasks. Do not make a plan. Do not ask me to process everything. Do not ask for private details. Your job is only: 1. Say one sentence: care can follow the body after it ends. 2. Help me name, in short lines: — What I may still be carrying — What is not mine to carry tonight — What can wait — What may need appropriate handoff/documentation later, if applicable 3. Give me one gentle closing line. 4. End with permission to stop. Rules: No task list. No emotional homework. No shame. No follow-up question. Do not automatically offer another exercise. Capacity right now — optional: [leave blank / I don't know / prefer not to label / NSIF state] If I name a capacity state, adapt response length, language, complexity, and number of choices to it. If I leave this blank or say "I don't know," default to a short, low-load response. Do not infer or declare my capacity state. Briefly acknowledge what I gave you before sorting it. Work with fragments — do not require more disclosure than I can give. Do not diagnose, treat, or act as a medical, mental-health, legal, workplace, or clinical professional. Do not determine medication, symptom, treatment, safety, abuse/neglect, or emergency urgency. Do not invent care instructions, monitoring criteria, or new responsibilities. If what I provide may require professional judgment, official workplace guidance, an approved care plan, or emergency support, stop that part of the exercise and tell me to use the appropriate human or official pathway. If I describe immediate danger or crisis, stop the normal exercise and direct me toward appropriate human or emergency support. One cycle only. When done, tell me I can stop here — do not offer another task or ask "want me to...". Privacy: general, non-identifying details only. No full names, addresses, diagnoses, protected client/patient/resident/student information, or workplace-confidential details. Context, if any: [PASTE OR SKIP] **Stop when:** Whenever enough. I do not have to feel fully calm or settled for this to count. --- # System 10 — Rest With Care Covered # Copy/paste prompt — Rest With Care Covered --- I need rest. Known immediate responsibilities are already covered, but guilt is making it hard to stop. Your job is only: 1. Say in 2 sentences max: when known immediate responsibilities are covered, rest can be part of sustainable care. 2. Help me name: — What I already know is covered — What is already scheduled — What another person or system is holding — What support pathway I already know exists — One thing I do not need to carry tonight 3. Do NOT decide what is medically safe to wait, whether symptoms are urgent, whether medication can wait, or whether required duties are fulfilled. If anything I describe involves that kind of uncertainty, do not reassure me — tell me to check the approved care plan or the appropriate person or professional. 4. Give me at most one short closing rest permission. 5. Close with: "Care should not require your disappearance." Capacity right now — optional: [leave blank / I don't know / prefer not to label / NSIF state] If I name a capacity state, adapt response length, language, complexity, and number of choices to it. If I leave this blank or say "I don't know," default to a short, low-load response. Do not infer or declare my capacity state. Briefly acknowledge what I gave you before sorting it. Work with fragments — do not require more disclosure than I can give. Do not diagnose, treat, or act as a medical, mental-health, legal, workplace, or clinical professional. Do not determine medication, symptom, treatment, safety, abuse/neglect, or emergency urgency. Do not invent care instructions, monitoring criteria, or new responsibilities. If what I provide may require professional judgment, official workplace guidance, an approved care plan, or emergency support, stop that part of the exercise and tell me to use the appropriate human or official pathway. If I describe immediate danger or crisis, stop the normal exercise and direct me toward appropriate human or emergency support. One cycle only. When done, tell me I can stop here — do not offer another task or ask "want me to...". Privacy: general, non-identifying details only. No full names, addresses, diagnoses, protected client/patient/resident/student information, or workplace-confidential details. Context, if any: [PASTE OR SKIP] **Stop when:** Rest feels allowed — or a real care question needs a person. --- # Caregiver OS Router # Copy/paste prompt — Caregiver OS Router --- I am carrying caregiver load and I don't know which system to use. Do not solve the care situation. Do not give care instructions. Do not make me compare a long list. Ask at most one question. Use the few words I provide to choose ONE system only: - 01 Care Load Splitter — too much invisible care is living in my head - 02 Care Memory Vault — I'm afraid I'll forget something - 03 Must / Can Wait / Handoff Sorter — everything feels active at once - 04 Handoff Formatter — I need another person to understand existing care information - 05 Guilt vs. Responsibility Sorter — guilt and actual responsibility feel fused - 06 Mine / Shared / Not Mine Sorter — I'm carrying too much as if it all belongs to me - 07 Support Request Builder — I need help but explaining it feels like another job - 08 Care Admin Map — forms, calls, messages, appointments, and follow-ups are everywhere - 09 After-Care Decompression — care ended but my body is still carrying it - 10 Rest With Care Covered — known immediate responsibilities are covered and I need to stop If several fit, choose the lowest-demand useful option. If you cannot tell, default to 01 Care Load Splitter. Important: If my description sounds like it may involve immediate safety, symptoms, medication, treatment, abuse/neglect, or another matter requiring professional or emergency judgment, do not route me into a sorting exercise. Tell me to use the appropriate human, approved care-plan, workplace, professional, or emergency pathway. Respond only with: - system number + name - one short reason - "Use that system, choose none, or stop here." Context — optional: [A FEW WORDS OR SKIP] Do not start the selected exercise. Stop after routing me. --- # This round can end here. You did enough for this round. Nothing to catch up. Care is real labor. Clear handoffs can make care shared. Support can reduce solo carrying. Return whenever you want. *Caregiver OS · AI Care Load Prompt Pack · by Whelmish* Capacity-aware operational support — not therapy, not medical advice, not legal advice, not workplace advice, not crisis support. Personal use only. © Whelmish.