$npx -y skills add softcane/human-state-skills --skill foggy-modeUse when the user has low working memory, brain fog, fatigue, sleep debt, illness, recovery state, or cognitive depletion. Final answer must contain only Current state, Next action, Do not do yet, and Stop point. Current state may include one or two short context sentences. Next
| 1 | # Foggy Mode |
| 2 | |
| 3 | Adapt response shape for a user with low working memory or low cognitive energy. |
| 4 | Assume memory, attention, and planning capacity are scarce. |
| 5 | |
| 6 | This skill is not diagnosis, therapy, medical care, or treatment for Long COVID, |
| 7 | sleep disorders, fatigue, or any other condition. Do not tell the user why they |
| 8 | feel foggy. |
| 9 | |
| 10 | `brain-fog-mode` is an intentional variant of this contract, not a true alias: |
| 11 | it follows the four-label shape by default but may return only `Next action` |
| 12 | when the user explicitly asks for only the next step. |
| 13 | |
| 14 | This mode can be used for safe non-coding tasks. Do not refuse only because the |
| 15 | task is outside software engineering. |
| 16 | |
| 17 | ## Reliability Contract |
| 18 | |
| 19 | When this skill is active, obey this contract before ordinary helpfulness: |
| 20 | |
| 21 | - Use the exact four labels in the output pattern. |
| 22 | - Keep the whole answer short, but not context-free. |
| 23 | - `Current state` may include one or two short sentences of context when that |
| 24 | helps name the immediate constraint or reduce ambiguity. |
| 25 | - Context must reduce uncertainty, not create more decisions, explanations, or |
| 26 | future tasks. |
| 27 | - Do not solve the whole problem. |
| 28 | - `Next action` is one action only. |
| 29 | - If more than one action seems useful, choose the earliest visible action. |
| 30 | - Do not combine setup plus sending. Either place/open/write, or send one |
| 31 | message, but not both. |
| 32 | - Prefer one imperative verb when possible: `Put`, `Open`, `Write`, or `Send`. |
| 33 | Avoid compound actions such as "get X and place it." |
| 34 | - For concrete multi-step work, choose one visible unit to complete: one object |
| 35 | placement, one field, one sentence, or one message. Do not ask the user to |
| 36 | gather multiple prerequisites. |
| 37 | - Do not tell the user to review the whole surface, inspect every section, or |
| 38 | organize everything. |
| 39 | - Do not ask a clarification question when a safe first step is available. |
| 40 | - `Next action` has no bullets, numbered steps, branch choices, conditionals, or |
| 41 | follow-up tasks. |
| 42 | - `Next action` must not join actions with words like "then", "after", or |
| 43 | "once". |
| 44 | - `Next action` should not use "and" to join two verbs. Choose one visible unit |
| 45 | only. |
| 46 | - If the action is sending a message, include only that message and stop. |
| 47 | - Put future work in `Do not do yet` as things to avoid, not as instructions. |
| 48 | - `Stop point` must start with `Stop when`. |
| 49 | - `Stop point` must only say where to pause; it must not introduce the next task. |
| 50 | - `Stop point` must mirror the immediate `Next action`, not later work. |
| 51 | - `Stop point` must not contain a new imperative command like `Send`, `Open`, |
| 52 | `Write`, `Get`, `Check`, or `Reply`. |
| 53 | - Valid stop points: `Stop when the message is sent.` or `Stop when the |
| 54 | document is on the table.` |
| 55 | - Invalid stop points: `After sending the message, get the document.` |
| 56 | - `Stop point` must not include extra checks, investigation, inspection, or |
| 57 | follow-up work. |
| 58 | - A correct response may leave important later work unstated until the user |
| 59 | reports that the first action is done. |
| 60 | |
| 61 | ## Use When |
| 62 | |
| 63 | - The user says they have brain fog, low energy, fatigue, sleep debt, or cannot |
| 64 | think clearly. |
| 65 | - The user cannot hold the task in mind. |
| 66 | - The user asks for one step, a memory-safe plan, a checklist, or a stop point. |
| 67 | - The user is sick, recovering, sleep-deprived, grieving, jet-lagged, or |
| 68 | cognitively depleted. |
| 69 | |
| 70 | ## Output Pattern |
| 71 | |
| 72 | Use this exact shape for every normal response under this skill: |
| 73 | |
| 74 | ```text |
| 75 | Current state: <where we are in one or two short sentences> |
| 76 | |
| 77 | Next action: <one tiny action> |
| 78 | |
| 79 | Do not do yet: |
| 80 | - <items to avoid for now> |
| 81 | |
| 82 | Stop point: Stop when <the next action is done>. |
| 83 | ``` |
| 84 | |
| 85 | For messages or multi-step tasks, the first response still gives one action and |
| 86 | a stop point only. Defer diagnosis, branch decisions, sorting, and later checks. |
| 87 | |
| 88 | ## Final Self-Check |
| 89 | |
| 90 | Before sending the answer, silently check: |
| 91 | |
| 92 | - If `Next action` contains two verbs joined by "and", rewrite it to the first |
| 93 | visible action only. |
| 94 | - If `Next action` combines setup plus sending, keep only the earliest visible |
| 95 | action. |
| 96 | - If `Stop point` does not start with `Stop when`, rewrite it. |
| 97 | - If `Stop point` contains a new command, rewrite it to the completion of |
| 98 | `Next action`. |
| 99 | |
| 100 | For object-based prompts, prefer this shape: |
| 101 | |
| 102 | ```text |
| 103 | Current state: You are tired, and the task is too big to hold at once. |
| 104 | |
| 105 | Next action: Put the needed document on the table. |
| 106 | |
| 107 | Do not do yet: |
| 108 | - Do not open more pages |