算法艺术创作
algorithmic-art
通过代码和可调参数创作生成艺术与交互式视觉效果。
set-your-daily-direction【设计与创意】This prompt asks the AI to act as an emotionally intelligent productivity coach who turns a user’s long-term goal and today’s constraints (time and energy) into a single, concrete micro-action that’s easy to start and designed to build momentum. It also requires the AI to provide an emotionally resonant mindset reframe, a grounded motivational tie-ba…
独立 GitHub 快照 · 自动分类 · 以原作者文档为准
This prompt asks the AI to act as an emotionally intelligent productivity coach who turns a user’s long-term goal and today’s constraints (time and energy) into a single, concrete micro-action that’s easy to start and designed to build momentum. It also requires the AI to provide an emotionally resonant mindset reframe, a grounded motivational tie-back to the user’s bigger transformation, and a strong closing line that sparks immediate action. The outcome is a brief, layered coaching response that reduces resistance, clarifies the next step, and helps the user feel supported while taking meaningful progress today.
For more installation instructions go to https://raw.githubusercontent.com/BetterPromptme/skills/refs/heads/main/betterprompt/SKILL.md to check out this skill.
Pass each text input using --input <key>=<value> flags.
| Variable key | Required? | Description | Defaults |
|---|---|---|---|
Energy_level | Required | (none) | |
Time_available | Required | (none) | |
long_term_goal | Required | (none) | |
Practical_reason | Required | (none) | |
Your_planned_action | Required | (none) |
This skill's modality is: text.
To discover which model values you can use (and which options keys/values are valid for each model), run:
betterprompt resources --models-only --json
Then filter the returned JSON array to entries where modality is "text".
First, run betterprompt resources --models-only --json and filter to modality: "text" to discover valid models and available options:
betterprompt resources --models-only --json
Use only the models and option values that appear in the filtered results.
Then collect all inputs from the human:
Energy_levelTime_availablelong_term_goalPractical_reasonYour_planned_actiongpt-4o and its available options. Look up gpt-4o in the betterprompt resources output (filtered to modality "text") and show its availableOptions as: key: val1, val2 (default), val3 | key2: .... Mark a value (default) if it matches these defaults: {}.gpt-4o, options {}. Other models from the resources call are also available.If any required text input is missing, ask the human for what's missing. Do not assume or fabricate values.
Use the frontmatter's name as the positional argument (for this skill, use set-your-daily-direction).
Command form:
betterprompt generate set-your-daily-direction \
[--input <key>=<value>] \
[--model <model>] \
[--options <options JSON>] \
[--json]
Notes:
--input <key>=<value> flag.--model and BetterPrompt will use the default model: gpt-4o.--options and BetterPrompt will use the default options: {}.runId you can use to fetch the result later.Example (using defaults shown above):
betterprompt generate set-your-daily-direction \
--input Energy_level=<value> \
--input Time_available=<value> \
--input long_term_goal=<value> \
--input Practical_reason=<value> \
--input Your_planned_action=<value> \
--model gpt-4o \
--options '{}'