一键导入
mood-to-motion
Turn procrastination into regulated action with an interview-driven emotion regulation prescription.
用 Codex 或 Claude 帮你安装 复制这段 Prompt,粘贴到 Codex、Claude 或其他助手里,让它检查 Skill 页面并帮你完成安装。
菜单
Turn procrastination into regulated action with an interview-driven emotion regulation prescription.
用 Codex 或 Claude 帮你安装 复制这段 Prompt,粘贴到 Codex、Claude 或其他助手里,让它检查 Skill 页面并帮你完成安装。
Learn and refine my writing voice through conversation.
Distill a book, method, or framework into an agent skill (Claude Code or any compatible agent) that follows a structured template (numbered framework sections, application tables, usage patterns, ethical boundaries, references/ deep-dives). Grounds every claim in verifiable sources — cites pages, URLs, or quotes; web-searches when the source isn't in front of the agent; drops anything that can't be cited. Uses one saved multi-agent workflow for extraction, coherent synthesis, reference expansion, and targeted review when authorized, with a portable direct-Agent fallback. Use when the user wants to turn a book / method / framework / paper / talk / course into a skill, or says "distill this", "make a skill from this book", "turn X into a skill", "book to skill", "source to skill".
Build and audit founder demand campaigns using Daniel Priestley’s Oversubscribed method, creating more qualified demand than honest delivery capacity before releasing an offer. Use when the user mentions “create more demand,” “launch oversubscribed,” “build a waitlist,” or asks how to generate excess demand for an offer. Also trigger when planning a launch, reviewing weak demand, measuring release readiness, or deciding how and when to release limited availability.
Facilitate an Inquiry through the canonical First Principles Thinking approach.
Solution Options is the second phase after problem-firewall: generate, stress-test, and evaluate solutions from a clear Problem Frame. Use when the user has a plain desired outcome, obstacle, and root-cause condition and wants to explore solution options, evaluate tradeoffs, run a premortem, avoid binary thinking, define decision criteria, or choose what to do next. For vague, symptom-level, or solution-first problems, use problem-firewall first.
Use when the user wants help solving something but the problem is vague, symptom-level, or framed as a premature solution. Enforce a problem-solution firewall: clarify the desired outcome, obstacles, root-cause condition, and user-owned plain-language Problem Frame before moving to solution methods.
基于 SOC 职业分类
| name | mood-to-motion |
| description | Turn procrastination into regulated action with an interview-driven emotion regulation prescription. |
| argument-hint | [optional: task or situation being avoided] |
| disable-model-invocation | true |
Run an interview that treats procrastination as maladaptive emotion regulation: the user avoids a task because avoidance repairs mood now. Replace that avoidance loop with the smallest evidence-backed regulation protocol that creates motion.
Do not lecture. Do not build a productivity system. Interview until the gates are satisfied, then prescribe. Ask one question at a time unless the user explicitly wants a written worksheet.
Use sources.md when writing the final cited section, when the user asks for the evidence, or when you need to distinguish technique claims.
The procrastination loop:
Task -> aversive emotion -> avoidance -> short-term relief -> reinforced avoidance
The Mood-to-Motion loop:
Task -> aversive emotion -> regulate to action threshold -> tiny approach action -> immediate reinforcement -> reinforced approach
The goal is not emotional comfort. The goal is enough regulation to act while some discomfort remains.
If the user already named the task, use it. Otherwise ask:
What task or situation are you avoiding?
Then ask only what is missing:
Completion criterion: you can state the loop as: "When facing <task>, you feel pulled to <avoidance behavior>, which gives short-term relief but creates <cost>. Today's visible progress is <finish>." If any bracket is vague, ask one more clarifying question before moving on.
The diagnostic fails if the wrong emotion is identified. Do not accept the user's first label automatically; test it against what happens right before avoidance.
Ask:
What feeling are you trying not to feel when you avoid this?
If the user is unsure, offer this menu and ask them to pick the strongest one:
Then run the driver lock:
<driver> — thoughts, body sensations, or the moment you switch to avoidance?"<driver> were reduced by half, would starting become easier?"Use this disambiguation map when labels conflict:
| Looks like | Lock it as | If the strongest evidence is |
|---|---|---|
| "Too much" | Overwhelm | The whole task feels present at once; relief comes from not looking at its size. |
| "I don't know how" | Uncertainty | The next step is unclear; relief comes from avoiding decisions or information gaps. |
| "What if this goes badly?" | Anxiety or fear | A future threat dominates; relief comes from escaping possible consequences. |
| "This says something about me" | Shame | Self-judgment or exposure dominates; relief comes from hiding from evaluation. |
| "This is dull" | Boredom | Understimulation dominates; relief comes from novelty, stimulation, or entertainment. |
| "I shouldn't have to" | Resentment | Obligation or unfairness dominates; relief comes from reclaiming control by refusing. |
| "It has to be good" | Perfection pressure | Evaluation standards block the first rep; relief comes from not producing a flawed version. |
| "I can't make myself" | Fatigue or depletion | Low energy/body state dominates even when the next action is clear and low-threat. |
| "I keep checking something else" | Distraction craving | A competing cue or reward hijacks attention; relief comes from the alternate stimulus. |
Completion criterion: the driver is locked only when the label, evidence, relief test, and counterfactual agree. One primary driver is named, its intensity is scored, and the short-term relief of avoidance is clear. If multiple drivers remain plausible, do not proceed; ask one contrast question: "Which one, if reduced by half, would make starting easiest: <driver A> or <driver B>?"
Use this map. Pick one primary technique and one support technique.
| Driver | Primary regulation | Support technique | Prescription logic |
|---|---|---|---|
| Overwhelm | Problem-solving reduction | Reappraisal | Cut the task to the first visible slice so the nervous system stops treating the whole project as now. |
| Uncertainty | Next-action clarification | Information boundary | Convert ambiguity into one physical action; block open-ended research. |
| Anxiety or fear | Slow breathing | Reappraisal + tiny exposure | Lower physiological arousal, then approach the feared task in a safe small dose. |
| Shame | Self-compassion | Repair action | Reduce self-attack so the user can re-engage without needing to feel worthy first. |
| Boredom | Reward pairing | Short timer | Create near reward and feedback because the distant outcome has weak pull. |
| Resentment | Autonomy restoration | Good-enough standard | Restore choice over how, when, or how much; reduce reactance. |
| Perfection pressure | Ugly first rep | Reappraisal | Lower evaluation threat by making the first version private, bad, or disposable. |
| Fatigue or depletion | Physiological reset | Minimum viable action | Regulate the body first, then shrink the ask below the energy barrier. |
| Distraction craving | Situation modification | Urge delay | Remove the cue or add friction; delay the competing reward long enough to start. |
Completion criterion: the chosen technique directly targets the named driver, not a generic productivity preference. If no technique clearly matches, ask one follow-up: "What makes starting feel emotionally expensive: size, uncertainty, threat, shame, boredom, resentment, energy, or temptation?"
Choose the start size from intensity:
Define the first action as physical and visible. It must create evidence of approach: an opened file, a written sentence, a sent message, a selected source, a moved object, a timer started, or a checklist item completed.
Completion criterion: the first action can be started now, is smaller than the user's resistance, and does not require motivation, a clear mood, or extra planning.
Output a prescription in this exact shape:
## Mood-to-Motion Prescription
Avoidance loop:
When I face <task>, I feel <driver/intensity>. Avoiding through <avoidance behavior> gives short-term relief, but it costs <cost>.
Driver lock:
I am treating this as <driver> because <evidence>, avoidance first relieves <relieved feeling>, and reducing <driver> would make starting easier.
Regulation target:
I am not trying to feel perfect. I am regulating from <intensity>/10 to action threshold.
Primary technique:
<technique> — <specific instruction for this task>
Support technique:
<technique> — <specific instruction for this task>
First approach action:
[ ] <30-second / 2-minute / 5-minute visible action>
Environment move:
<remove cue, open tool, prepare surface, block distraction, or lower setup friction>
Reinforcement:
After the action, <immediate reward/checkmark/visible progress statement>.
If-then:
If I still avoid, then I will <smaller fallback action>.
Cited evidence:
- Why this targets procrastination: <Sirois & Pychyl, 2013; Steel, 2007; optional Tice & Baumeister, 1997 if emphasizing short-term relief / long-term cost>.
- Why this regulation family fits: <Gross, 1998/2002; Webb et al., 2012; Aldao et al., 2010>.
- Technique-specific support: <slow breathing / self-compassion / problem-solving citation if used>.
After the prescription, tell the user:
Do the first approach action now, then come back with: done / blocked / avoided.
Completion criterion: the prescription contains a named avoidance loop, a driver lock, a regulation target, one primary technique, one support technique, one physical first action, one environment move, one reinforcement, one fallback, and compact cited evidence.
If the user returns with done, ask:
Intensity now: lower, same, or higher?
Then either prescribe the next approach action or a clean stop.
If the user returns with blocked, ask:
What physically or practically blocked the action?
Remove that blocker and shrink the action.
If the user returns with avoided, say:
Good data, not failure. What happened in the moment you were supposed to start?
Rediagnose once. If they avoid twice, stop analysis and prescribe a 30-second environment-first action.
Always include compact cited evidence in the final prescription. If the user asks for sources or challenges a claim, provide the fuller cited section from sources.md with links and explain which source supports which part of the prescription.