meta-kid-project-planner

star 4.3k

Use this meta-skill instead of answering directly when a child or their guardian wants to plan a school project, science fair entry, hobby kit, or kid-sized creative venture (volcano model, bug-watching YouTube channel, magnet maze, model rocket). The skill assesses feasibility against the child's age band, builds an age-appropriate step plan, lists materials with budget substitutes, surfaces safety considerations, and produces a parent-facing learning-objective summary so the guardian can supervise meaningfully. Do not use it for adult craft projects, generic art prompts, generic school-project explanations, or unsafe projects. Refuses inappropriate or unsafe projects.

opensquilla By opensquilla schedule Updated 6/12/2026
play_arrow Run Skill in Manus View GitHub

name: meta-kid-project-planner description: "Use this meta-skill instead of answering directly when a child or their guardian wants to plan a school project, science fair entry, hobby kit, or kid-sized creative venture (volcano model, bug-watching YouTube channel, magnet maze, model rocket). The skill assesses feasibility against the child's age band, builds an age-appropriate step plan, lists materials with budget substitutes, surfaces safety considerations, and produces a parent-facing learning-objective summary so the guardian can supervise meaningfully. Refuses inappropriate or unsafe projects." kind: meta meta_priority: 60 always: false final_text_mode: "step:project_pack_audit" request_template: outcome: "Age-appropriate project plan with materials, safety notes, and guardian guidance." outcome_zh: "适合孩子年龄的项目计划，包含材料、安全提示和家长指导。" outcome_en: "Age-appropriate project plan with materials, safety notes, and guardian guidance." fields: - name: project_topic label_zh: "项目主题" label_en: "Project topic" required: true - name: age_band label_zh: "年龄段" label_en: "Age band" required: false default: "early grade" default_zh: "小学低年级" default_en: "early grade" - name: deadline_days label_zh: "截止天数" label_en: "Days until due" required: false - name: budget_or_material_constraints label_zh: "预算或材料限制" label_en: "Budget or material constraints" required: false - name: audience label_zh: "受众" label_en: "Audience" required: false default: "child plus guardian" default_zh: "孩子和监护人" default_en: "child plus guardian" - name: language label_zh: "输出语言" label_en: "Output language" required: false default: "match the user's language" default_zh: "跟随用户语言" default_en: "match the user's language" assumptions: - "Refuse unsafe or age-inappropriate projects." - "Default to modest materials and light guardian supervision when unspecified." assumptions_zh: - "拒绝不安全或不适合年龄的项目。" - "未说明时，默认使用适度材料并安排轻量家长监督。" assumptions_en: - "Refuse unsafe or age-inappropriate projects." - "Default to modest materials and light guardian supervision when unspecified." output_contract: append_to_final_text: false required_sections: - "Feasibility verdict" - "Step-by-step plan" - "Materials and substitutions" - "Safety notes" - "Guardian learning objectives" assumptions: - "Age band and supervision level may be defaulted when absent." unverified: - "Local material availability and school-specific rubric details." artifacts: - name: "project_pack" required: false eval_prompts:

  • name: "kid-project-baseline" prompt: "Plan a safe age-appropriate science fair project for an elementary-school child with a small budget." rubric:
    • "Feasibility verdict"
    • "Step-by-step plan"
    • "Materials and substitutions"
    • "Safety notes"
    • "Guardian learning objectives" preference_keys:
  • preferred_language
  • guardian_supervision_level policy_tags:
  • child-safety
  • age-appropriate triggers:
  • "school project"
  • "science fair"
  • "kid science"
  • "孩子做项目"
  • "做一个手工"
  • "科学课作业"
  • "help my kid build"
  • "我要做火山"
  • "child diy project"
  • "课外动手项目" provenance: origin: opensquilla-original license: Apache-2.0 metadata: opensquilla: risk: medium capabilities: [network, filesystem-write] clawhub_top100_composition:
    • skill: "Multi Search Engine" local_skill: multi-search-engine rank_source: "Top ClawHub Skills downloads top100, 2026-05-28" rank: 11 role: "Find safe, age-appropriate how-to references and material alternatives."
    • skill: "Weather" local_skill: weather rank_source: "Top ClawHub Skills downloads top100, 2026-05-28" role: "Plan outdoor child projects around realistic weather constraints."
    • skill: "Deep Researcher / deep research family" local_skill: deep-research rank_source: "ClawHub research-skill family, verified via current search results" role: "Add extra safety and feasibility research when the project is more complex."
    • skill: "PowerPoint / PPTX" local_skill: pptx rank_source: "Top ClawHub Skills downloads top100, 2026-05-28" role: "Produce kid-facing printable step cards or a simple presentation when requested."

composition: steps: - id: preferences label: "偏好提取" label_en: "Preference extraction" kind: llm_chat with: system: "You extract kid-project preferences. Return only the requested contract. Refuse to plan projects that are clearly unsafe (firearms, fireworks, drugs, sharp-weapon making, etc.) by setting PROJECT_SAFE: no." task: | Extract the kid-project brief.

      User request:
      {{ inputs.user_message | xml_escape | truncate(1600) }}

      Clarification policy:
      - If the request already includes a project topic, child age or age
        band, and deadline or rough time window, set
        NEEDS_CLARIFICATION: no and MISSING_FIELDS: none.
      - Budget, exact presentation format, exact weather, and exact
        sunshine direction can remain explicitly unknown assumptions; do
        not block on them and do not invent precise values.
      - For common school projects, proceed with explicit assumptions or
        UNKNOWN markers instead of asking a form unless safety or the core
        topic is unclear.

      Return exactly:
      TOPIC: <short project description>
      AGE_BAND: <PRE_K|EARLY_GRADE|TWEEN|TEEN|UNKNOWN>
      DEADLINE_DAYS: <integer days until due, or UNKNOWN>
      BUDGET_BAND: <SHOESTRING|MODEST|COMFORTABLE|UNKNOWN>
      PARENT_SUPERVISION: <SOLO|LIGHT|HANDS_ON|UNKNOWN>
      LANGUAGE: <en|zh|mixed>
      PROJECT_SAFE: <yes|no>
      UNSAFE_REASON: <one phrase if PROJECT_SAFE is no, else none>
      NEEDS_CLARIFICATION: <yes|no>
      MISSING_FIELDS:
        - <topic|age_band|deadline|none>
      ASSUMPTIONS:
        - <assumption>
- id: project_clarify
  label: "项目澄清"
  label_en: "Project clarification"
  kind: user_input
  depends_on: [preferences]
  when: "'NEEDS_CLARIFICATION: yes' in outputs.preferences and 'PROJECT_SAFE: yes' in outputs.preferences and 'MISSING_FIELDS:\n  - none' not in outputs.preferences"
  clarify:
    mode: form
    intro: |
      再确认几件事，然后给你完整的项目规划（含家长版） / A few details and I'll build the kid + parent pack.
    intro_zh: "再确认几件事，然后给你完整的项目规划（含家长版）。"
    intro_en: "A few details and I will build the kid plus guardian project pack."
    nl_extract: true
    fields:
      - name: topic
        type: string
        required: true
        prompt: "项目主题（如：做一座火山模型）/ Project topic"
        prompt_zh: "项目主题（例如：做一座火山模型）"
        prompt_en: "Project topic"
        max_chars: 200
      - name: age_band
        type: enum
        required: true
        choices: [PRE_K, EARLY_GRADE, TWEEN, TEEN]
        prompt: "孩子年龄段 / Child age band (PRE_K = 3-5, EARLY_GRADE = 6-9, TWEEN = 10-12, TEEN = 13-17)"
        prompt_zh: "孩子年龄段（PRE_K=3-5 岁，EARLY_GRADE=6-9 岁，TWEEN=10-12 岁，TEEN=13-17 岁）"
        prompt_en: "Child age band (PRE_K = 3-5, EARLY_GRADE = 6-9, TWEEN = 10-12, TEEN = 13-17)"
      - name: deadline_days
        type: int
        min: 0
        max: 365
        default: 14
        prompt: "几天后要交（0 = 今天，14 = 两周）/ Days until due"
        prompt_zh: "几天后要交（0 = 今天，14 = 两周）"
        prompt_en: "Days until due"
      - name: budget_band
        type: enum
        choices: [SHOESTRING, MODEST, COMFORTABLE]
        default: MODEST
        prompt: "预算 / Budget"
        prompt_zh: "预算"
        prompt_en: "Budget"
      - name: parent_supervision
        type: enum
        choices: [SOLO, LIGHT, HANDS_ON]
        default: LIGHT
        prompt: "家长参与程度（SOLO 几乎不参与；LIGHT 偶尔帮一下；HANDS_ON 全程在旁）/ Parent supervision"
        prompt_zh: "家长参与程度（SOLO 几乎不参与；LIGHT 偶尔帮一下；HANDS_ON 全程在旁）"
        prompt_en: "Parent supervision (SOLO = mostly independent; LIGHT = occasional help; HANDS_ON = close supervision)"
      - name: language
        type: enum
        choices: [en, zh, mixed]
        default: mixed
        prompt: "输出语言 / Language"
        prompt_zh: "输出语言"
        prompt_en: "Output language"
    cancel_keywords: ["算了", "换个项目", "cancel", "stop"]
    timeout_hours: 48
- id: feasibility
  label: "可行性"
  label_en: "Feasibility"
  kind: llm_classify
  depends_on: [preferences, project_clarify]
  output_choices:
    - STRAIGHTFORWARD
    - NEEDS_ADULT_HELP
    - NEEDS_SHOPPING
    - SAFETY_REVIEW_REQUIRED
    - INAPPROPRIATE
  with:
    text: |
      Classify project feasibility for this child.

      Topic: from preferences / clarify.
      Preferences:
      {{ outputs.get('preferences', '') | truncate(800) }}
      Clarification:
      {{ inputs.get('collected', {}).get('project_clarify', {}) | tojson }}

      Decision rules:
      - INAPPROPRIATE: project involves weapons, fire without
        supervision possibility, drugs, harmful chemistry,
        self-harm-adjacent themes, or other clearly unsafe topics
        for any minor.
      - SAFETY_REVIEW_REQUIRED: project involves heat (small stove,
        soldering iron), sharp blades (X-Acto knives), electronics
        with mains voltage, or moderately reactive chemistry. Adult
        must be present.
      - NEEDS_SHOPPING: project requires materials the household
        likely does not have (specific kit, model rocket motor,
        specialty paint).
      - NEEDS_ADULT_HELP: project is age-appropriate but a step
        requires hands the child does not yet have (cutting balsa
        wood for a 6-year-old, threading a needle for a 4-year-old).
      - STRAIGHTFORWARD: child can complete the bulk of the project
        with the declared PARENT_SUPERVISION level.
- id: redirect_unsafe
  label: "安全改写"
  label_en: "Safety rewrite"
  kind: llm_chat
  depends_on: [preferences, feasibility]
  when: "'PROJECT_SAFE: no' in outputs.preferences or outputs.feasibility == 'INAPPROPRIATE'"
  with:
    system: "You write a gentle, non-shaming redirect when a project topic is unsafe or inappropriate. Always offer 3 alternative project ideas that are in the same SPIRIT as the original (curiosity-driven, hands-on, age-appropriate)."
    task: |
      Topic the user asked for:
      {{ inputs.user_message | xml_escape | truncate(400) }}

      Unsafe reason (from preferences):
      {{ outputs.get('preferences', '') | truncate(400) }}

      Write:
      1. One sentence acknowledging the curiosity behind the
         original idea (do not lecture).
      2. One sentence explaining gently why this version isn't a
         good kid project (concrete, not vague).
      3. Three alternative project ideas that scratch a similar
         itch but are safe and age-appropriate. Each: one-line
         topic + one-line "what the kid will learn / make".

      Language: match preferences (default zh).
      End with a single line:
      UNSAFE_REDIRECT: yes
- id: recall_past_projects
  label: "项目召回"
  label_en: "Project recall"
  kind: tool_call
  tool: memory_search
  tool_allowlist: [memory_search]
  depends_on: [feasibility, project_clarify]
  when: "outputs.feasibility != 'INAPPROPRIATE' and 'PROJECT_SAFE: yes' in outputs.preferences"
  on_failure: recall_past_projects_fallback
  tool_args:
    query: "child project prior projects preferences constraints {{ outputs.get('preferences', '') | truncate(600) }} {{ inputs.user_message | xml_escape | truncate(600) }}"
    max_results: 6
    source: memory
- id: recall_past_projects_fallback
  label: "项目召回兜底"
  label_en: "Project recall fallback"
  kind: llm_chat
  with:
    system: "You produce a no-memory fallback note for child project planning."
    task: |
      No durable project memory was read. Continue using only the pasted
      child age, deadline, materials, budget, supervision, location, and
      project context. Do not mention runtime errors to the user.
- id: web_research
  label: "网页研究"
  label_en: "Web research"
  kind: skill_exec
  skill: multi-search-engine
  depends_on: [feasibility]
  when: "outputs.feasibility in ['STRAIGHTFORWARD', 'NEEDS_ADULT_HELP', 'NEEDS_SHOPPING', 'SAFETY_REVIEW_REQUIRED']"
  on_failure: web_research_fallback
  with:
    query: "{{ inputs.get('collected', {}).get('project_clarify', {}).get('topic', '') }} {{ outputs.get('preferences', '') | truncate(160) }} {{ inputs.user_message | xml_escape | truncate(160) }} kid science project step-by-step instructions safe"
    engines: [brave, tavily, duckduckgo]
    max_results: 8
- id: web_research_fallback
  label: "网页研究兜底"
  label_en: "Web research fallback"
  kind: llm_chat
  with:
    system: "You produce a no-web fallback note for child project planning."
    task: |
      Web research was not available. Extract the project topic, age,
      deadline, materials, budget, parent availability, location, light,
      weather-sensitive constraints, and safety needs only from the pasted
      request. Do not expose tool names, paths, stack traces, connector
      wording, or runtime failures.

      Request:
      {{ inputs.user_message | xml_escape | truncate(3500) }}
- id: weather_check
  label: "天气检查"
  label_en: "Weather check"
  kind: skill_exec
  skill: weather
  depends_on: [feasibility, project_clarify]
  when: "outputs.feasibility != 'INAPPROPRIATE' and ('outdoor' in (inputs.user_message | lower) or 'balcony' in (inputs.user_message | lower) or 'plant' in (inputs.user_message | lower) or 'garden' in (inputs.user_message | lower) or 'park' in (inputs.user_message | lower) or '户外' in inputs.user_message or '阳台' in inputs.user_message or '植物' in inputs.user_message or '豆芽' in inputs.user_message)"
  on_failure: weather_check_fallback
  with:
    location: "{{ inputs.user_message | xml_escape | truncate(60) }}"
    days: 7
- id: weather_check_fallback
  label: "天气兜底"
  label_en: "Weather fallback"
  kind: llm_chat
  with:
    system: "You produce a no-live-weather fallback note for child project planning."
    task: |
      Live weather was not verified. Continue using only the pasted
      location and the user's supplied light/outdoor/indoor context. Do not
      infer forecasts, temperature ranges, rainfall, balcony direction,
      sunshine hours, or season-specific claims. Do not mention tool
      failures.

      Request:
      {{ inputs.user_message | xml_escape | truncate(2000) }}
- id: project_fact_ledger
  label: "项目事实台账"
  label_en: "Project fact ledger"
  kind: llm_chat
  depends_on: [preferences, project_clarify, recall_past_projects, weather_check]
  when: "outputs.feasibility != 'INAPPROPRIATE' and 'PROJECT_SAFE: yes' in outputs.preferences"
  with:
    system: "You extract a strict source-fact ledger for a child project plan. You do not write the plan. You separate user-provided facts, durable memory facts, unknowns, and unsafe or unsupported inferences."
    task: |
      Build a strict project fact ledger from the user's request and any
      clarification payload. Durable memory is also a source of facts when
      the memory output clearly states child profile, prior projects,
      preferences, or parent availability. Ignore memory status prose,
      file inventory, workspace paths, and runtime/tool wording; extract
      only the actual remembered facts.

      User request:
      {{ inputs.user_message | xml_escape | truncate(3500) }}

      Clarification:
      {{ inputs.get('collected', {}).get('project_clarify', {}) | tojson | truncate(1200) }}

      Durable memory / past-project recall:
      {{ outputs.get('recall_past_projects', '') | truncate(1800) }}

      Weather result or fallback:
      {{ outputs.get('weather_check', '') | truncate(800) }}

      Return exactly:
      OUTPUT_LANGUAGE: <zh|en|mixed>
      PROVIDED_CHILD_CONTEXT:
        - <age, ability, child preferences, guardian availability, or UNKNOWN>
      PROVIDED_MEMORY_CONTEXT:
        - <remembered child profile, prior projects, parent constraints, lessons learned, or none>
      PROVIDED_PROJECT_CONTEXT:
        - <topic, school deadline/time window, school output, or UNKNOWN>
      PROVIDED_MATERIALS_BUDGET:
        - <available materials and budget exactly as supplied, or UNKNOWN>
      PROVIDED_LOCATION_LIGHT:
        - <location and light exactly as supplied, or UNKNOWN>
      VERIFIED_WEATHER:
        - <verified forecast if actually present in weather result, else none>
      UNKNOWN_DETAILS:
        - <exact date, balcony direction, exact weather, exact school format, etc.>
      FORBIDDEN_INFERENCES:
        - <details that must not appear as facts, e.g. exact calendar date,
          balcony faces south/east/west, temperature range, rain forecast,
          school rule, allergy, fake measurements, tasting/eating>

      Rules:
      - If durable memory says the child is a specific age, likes/dislikes
        an activity style, has prior projects, or has parent time limits,
        put those in PROVIDED_CHILD_CONTEXT and PROVIDED_MEMORY_CONTEXT.
        Do not mark them UNKNOWN.
      - If the current request asks not to repeat previous projects, list
        remembered prior projects in PROVIDED_MEMORY_CONTEXT so downstream
        steps can avoid them explicitly.
      - If the source says only "two weeks later", mark exact date UNKNOWN.
      - If the source says only "half-day sun", mark orientation and hours
        UNKNOWN. Preserve "half-day sun" exactly.
      - If weather was not actually verified, VERIFIED_WEATHER must be none.
      - Never invent sample measurements, dates, allergies, school rules,
        or local forecasts.
- id: deep_research
  label: "深度研究"
  label_en: "Deep research"
  kind: skill_exec
  skill: deep-research
  depends_on: [feasibility]
  when: "outputs.feasibility in ['SAFETY_REVIEW_REQUIRED', 'NEEDS_SHOPPING']"
  with:
    query: "{{ inputs.get('collected', {}).get('project_clarify', {}).get('topic', '') }} safe materials children"
    depth: "standard"
    max_rounds: 1
- id: outline_steps
  label: "步骤大纲"
  label_en: "Step outline"
  kind: llm_chat
  depends_on: [feasibility, web_research, recall_past_projects, weather_check, project_clarify, preferences]
  when: "outputs.feasibility != 'INAPPROPRIATE' and 'PROJECT_SAFE: yes' in outputs.preferences"
  with:
    system: "You break a project into kid-sized steps. Each step takes 5-20 minutes for an unhurried child. No step requires reading more than a paragraph. Use concrete, doable verbs."
    task: |
      Build a step-by-step plan grounded in the topic and research.

      Topic + context:
      {{ outputs.get('preferences', '') | truncate(500) }}
      {{ inputs.get('collected', {}).get('project_clarify', {}) | tojson }}
      If clarification fields are empty, extract topic, age, deadline,
      budget, materials, parent availability, location, and light from the
      user request:
      {{ inputs.user_message | xml_escape | truncate(2500) }}

      Past projects this child has done (avoid repeats; build on prior learning):
      {{ outputs.get('recall_past_projects', '') | truncate(600) }}

      Weather (for outdoor projects — pick a good day):
      {{ outputs.get('weather_check', '') | truncate(400) }}

      Web research:
      {{ outputs.get('web_research', '') | truncate(2500) }}

      Deep research (only present if research happened):
      {{ outputs.get('deep_research', '') | truncate(2000) }}

      Deadline: {{ inputs.get('collected', {}).get('project_clarify', {}).get('deadline_days', 14) }} days.
      Age band: {{ inputs.get('collected', {}).get('project_clarify', {}).get('age_band', 'EARLY_GRADE') }}.
      Parent supervision: {{ inputs.get('collected', {}).get('project_clarify', {}).get('parent_supervision', 'LIGHT') }}.

      Output a markdown numbered list. Each step has:
      - Title (one short sentence, kid-readable verb-first)
      - Time estimate (5-20 min)
      - Adult-needed: yes / no (be honest)
      - One supportive sentence ("You'll get to ...")

      Distribute the steps across the available days. If deadline is
      tight, mark which steps to skip or shorten.

      Language: match preferences (default mixed).
- id: material_list
  label: "材料清单"
  label_en: "Materials list"
  kind: llm_chat
  depends_on: [outline_steps, project_clarify, web_research]
  when: "outputs.feasibility != 'INAPPROPRIATE' and 'PROJECT_SAFE: yes' in outputs.preferences"
  with:
    system: "You list materials needed for a kid project. For each material, offer a SHOESTRING substitute the family likely has at home. Always be honest when a substitute genuinely won't work."
    task: |
      List materials needed.

      Step plan:
      {{ outputs.outline_steps | truncate(2500) }}

      Budget band: {{ inputs.get('collected', {}).get('project_clarify', {}).get('budget_band', 'MODEST') }}.

      Output a markdown table with columns:
      item | quantity | est_cost | shoestring_substitute |
      notes_if_substitute_changes_outcome.

      Add at the end a bullet line "Likely you have at home:" listing
      items the household probably already owns.
- id: safety_notes
  label: "安全提示"
  label_en: "Safety notes"
  kind: llm_chat
  depends_on: [feasibility, outline_steps, project_clarify]
  when: "outputs.feasibility != 'INAPPROPRIATE' and 'PROJECT_SAFE: yes' in outputs.preferences"
  with:
    system: "You surface safety considerations for a kid project. Be specific (not generic 'be careful'). Tailor to the age band. If feasibility says SAFETY_REVIEW_REQUIRED, the safety section is the most important part of the deliverable."
    task: |
      Surface safety notes specific to this project.

      Feasibility: {{ outputs.feasibility }}
      Age band: {{ inputs.get('collected', {}).get('project_clarify', {}).get('age_band', 'EARLY_GRADE') }}.
      Step plan (for reference):
      {{ outputs.outline_steps | truncate(2000) }}

      Output bullet points grouped under:
      ## ⚠️ Adult must be present for
      ## ✋ Stop and call an adult if
      ## 🧪 Materials to handle carefully

      Each bullet must reference a specific step from the plan (e.g.
      "Step 4 (mixing): vinegar can spray — wear safety glasses or
      old sunglasses").

      If feasibility is STRAIGHTFORWARD and the child is TWEEN+,
      keep this section short (3-5 bullets). If SAFETY_REVIEW_REQUIRED
      or younger child, be thorough (8-12 bullets).

      Language: match preferences (default mixed).
- id: learning_objectives
  label: "学习目标"
  label_en: "Learning goals"
  kind: llm_chat
  depends_on: [outline_steps, preferences, project_clarify]
  when: "outputs.feasibility != 'INAPPROPRIATE' and 'PROJECT_SAFE: yes' in outputs.preferences"
  with:
    system: "You write the parent-facing learning-objective section. This is what the GUARDIAN reads to know what their child is actually getting out of the project beyond the artifact."
    task: |
      Write the parent-facing learning objectives.

      Topic + step plan:
      {{ outputs.get('preferences', '') | truncate(400) }}
      {{ outputs.outline_steps | truncate(2500) }}

      Output:
      ## 👀 What your kid will actually learn

      3-5 bullets, each one concrete learning outcome grounded in a
      specific step. Avoid generic outcomes like "creativity" or
      "problem-solving" — name the specific concept (e.g. "How an
      acid-base reaction releases CO2 — they'll see the bubbling
      slowdown when vinegar runs out").

      ## 🧠 Conversation prompts during/after
      3 questions the parent can ask the child to deepen the
      learning. Each question must be open-ended.

      Language: match preferences (parent-facing — usually adult
      register).
- id: kid_deck
  label: "儿童卡片"
  label_en: "Child card"
  kind: skill_exec
  skill: pptx
  depends_on: [outline_steps, material_list, safety_notes, project_clarify, feasibility]
  when: "outputs.feasibility != 'INAPPROPRIATE' and 'PROJECT_SAFE: yes' in outputs.preferences and inputs.get('collected', {}).get('project_clarify', {}).get('parent_supervision', 'LIGHT') == 'HANDS_ON'"
  with:
    mode: create
    title: "🛠️ {{ inputs.get('collected', {}).get('project_clarify', {}).get('topic', 'Your project') }}"
    slides:
      - title: "What we're going to make"
        body: "{{ outputs.get('preferences', '') | truncate(400) }}"
      - title: "What you need (materials)"
        body: "{{ outputs.get('material_list', '') | truncate(800) }}"
      - title: "Step-by-step plan"
        body: "{{ outputs.get('outline_steps', '') | truncate(1200) }}"
      - title: "⚠️ Stop and call grown-up if"
        body: "{{ outputs.get('safety_notes', '') | truncate(600) }}"
    output_path: "kid_project_{{ inputs.get('collected', {}).get('project_clarify', {}).get('topic', 'untitled') | slugify }}.pptx"
- id: vocab_cards
  label: "词汇卡"
  label_en: "Vocabulary card"
  kind: llm_chat
  depends_on: [outline_steps, material_list, project_clarify, feasibility, preferences]
  when: "outputs.feasibility != 'INAPPROPRIATE' and 'PROJECT_SAFE: yes' in outputs.preferences and ('vocab' in (inputs.user_message | lower) or 'word card' in (inputs.user_message | lower) or 'bilingual' in (inputs.user_message | lower) or '英语' in inputs.user_message or '双语' in inputs.user_message or '单词' in inputs.user_message)"
  with:
    system: "You produce a small vocabulary card list grounded in the project content. Each card is age-appropriate."
    task: |
      Produce 6 vocab cards from the project content.

      Step plan:
      {{ outputs.outline_steps | truncate(2000) }}

      Materials:
      {{ outputs.material_list | truncate(800) }}

      Age band: {{ inputs.get('collected', {}).get('project_clarify', {}).get('age_band', 'EARLY_GRADE') }}.

      For each card:
      - Term (one English word + one Chinese gloss if the language
        preference is `mixed`)
      - Kid-friendly definition (one sentence)
      - Example sentence from the project ("In step 3, the vinegar
        ACID meets the baking soda BASE...")

      Output as a markdown numbered list.
- id: deliver_project_pack
  label: "项目包交付"
  label_en: "Project package delivery"
  kind: llm_chat
  depends_on:
    - preferences
    - feasibility
    - redirect_unsafe
    - outline_steps
    - material_list
    - safety_notes
    - learning_objectives
    - vocab_cards
    - kid_deck
    - recall_past_projects
    - weather_check
    - project_fact_ledger
    - project_clarify
  with:
    system: "You assemble the final project pack the user will read. Return the complete deliverable inline in chat. Do not create, save, export, attach, or point primarily to an artifact unless the user explicitly asked for a file export with words like PDF, file, export, attachment, or download. Treat requests for a printable worksheet, printable record sheet, or poster layout as print-ready markdown included inline. Never mention workflow, meta-skill, tool names, connector failures, workspace paths, or runtime details."
    task: |
      Assemble the final project pack.

      If the project was unsafe (redirect_unsafe ran), output ONLY
      the content of {{ outputs.get('redirect_unsafe', '') }} verbatim
      and end with PACK_DELIVERED: no_safety_redirect — nothing else.

      Otherwise, synthesize a concise, non-duplicative project pack.
      do not copy intermediate outputs verbatim when that would repeat
      safety text, parent instructions, or child instructions. Use the
      intermediate outputs as source material and rewrite them into one
      practical final answer.

      Original user request:
      {{ inputs.user_message | xml_escape | truncate(3500) }}

      Project fact ledger:
      {{ outputs.get('project_fact_ledger', '') | truncate(2500) }}

      Durable memory / past-project recall:
      {{ outputs.get('recall_past_projects', '') | truncate(1600) }}

      Source-constraint audit:
      - Treat the Project fact ledger as the source of truth. If an
        intermediate step conflicts with it, ignore the intermediate step.
      - Treat PROVIDED_MEMORY_CONTEXT in the fact ledger as user-relevant
        durable context. Preserve remembered age, child preferences, parent
        time limits, prior projects, and lessons learned unless the current
        request explicitly contradicts them.
      - If the fact ledger marks a detail UNKNOWN, leave it unknown or mark
        it as an assumption; do not fill it with common sense or current
        date/time.
      - Preserve every explicit user constraint before adding suggestions:
        age, deadline, location, available materials, budget,
        parent time, light/weather constraints, school deliverable, and
        requested output sections.
      - Do not replace user-provided materials with unrelated materials.
        Prefer the user's materials first, then list substitutes only as
        backups.
      - Do not invent calendar dates, school dates, temperature readings,
        weather forecasts, or live conditions. Use day numbers or say the
        exact date is unknown unless the user supplied one or live data was
        verified.
      - If the user gives only a relative deadline such as "two weeks
        later", do not convert it into a calendar date. Say "两周后 /
        relative deadline supplied; exact date not provided" and schedule
        by Day 1...Day 14.
      - Do not invent balcony direction, temperature ranges, sunshine
        hours, rain forecasts, school rules, allergies, or local weather.
        Keep weather/light claims to what the user supplied plus clearly
        labelled assumptions.
      - Do not prefill observation tables with fake measurements, fake
        dates, or predicted heights. For templates, leave measurement cells blank or as placeholders
        such as "__ cm" / "[记录]".
      - Do not suggest tasting or eating the experiment materials unless
        the user explicitly asks for an edible-food activity and safety has
        been reviewed. Observation projects should compare appearance,
        height, color, firmness by sight/touch if safe, and drawings.
      - Design a comparison experiment when it fits the project and stays
        within the user's materials, age, time budget, and safety limits.
        For observation projects, make the variable, control, and data table
        clear enough for a school presentation.
      - Prefer a clear comparison design for plant/observation projects:
        same seed, cup, water, and paper-towel conditions, with only one
        changed variable such as light exposure. If materials allow, use
        2-3 labelled groups; if not, make the single-group observation
        plan still presentation-ready.
      - "Printable" means a clean markdown table, worksheet block, or
        poster-board layout that the user can print from chat. Do not
        create or refer to PDFs, HTML files, downloads, attachments,
        local paths, generated artifacts, or workspace files unless the
        user explicitly asked for a file/PDF/export/download.
      - If the user asks for a beautiful or visually polished plan, make
        the inline markdown itself polished: a memorable title, a concise
        visual theme, color/palette suggestions, kid-facing labels,
        a drawing-heavy record sheet, and a parent-ready poster layout.
      - If the user asks to start with remembered constraints, include a
        section titled exactly "## Remembered constraints I used" near the
        top and list only facts found in PROVIDED_MEMORY_CONTEXT or the
        current request. Do not invent memory.

      Language and length:
      - Match the user's language. For Chinese requests, write Simplified
        Chinese throughout, including headings and child-facing text.
      - For Chinese requests, do not use English section headings such as
        "For You (the kid)" or "For the Grown-up".
      - Do not include vocabulary cards unless the user explicitly asked
        for vocab, word cards, bilingual support, 英语, 双语, or 单词.
      - For straightforward home/school projects, target 1800-3200 Chinese characters
        or an equivalent compact English length unless the user asks for a
        long worksheet.

      For Chinese safe projects, use this structure:

      # 🛠️ <Project Title>

      ## 先说假设
      State the age/deadline/material/weather assumptions and which facts
      are not live-verified.

      ## 项目设计
      Explain the child-friendly project question, final deliverable, and
      2-3 learning goals in plain language.

      ## 14 天计划
      Give an actionable schedule. Group days into phases when that is
      clearer than 14 long paragraphs, but preserve deadlines, daily
      observation habits, parent touchpoints, and what the child does.

      ## 材料和替代品
      Summarize the materials table and substitutions. Keep shopping and
      household alternatives practical.

      ## 安全和翻车点
      Include only the safety points that change behavior: stop-and-call
      conditions, allergy/toxin/sharp/hot/water/electricity risks where
      relevant, and 3 common failure modes with fixes.

      ## 数据记录和画图
      Include a simple observation template for date, height, leaf/color,
      water, light, and one child-friendly chart idea.

      ## 天气/光照调整
      Use live weather only if verified; otherwise state "live weather not
      verified / 实时天气未核验" and give safe indoor/low-light alternatives.

      ## 最后展示怎么讲
      Provide a 60-second child script and 3 likely teacher questions with
      simple answers.

      ## 家长每晚 20 分钟
      Summarize what the adult checks each night and which steps require
      hands-on help.

      ## 还缺哪些实时信息
      List only genuinely missing information that would improve the plan,
      without asking the user to confirm before using the current answer.

      For English safe projects, use equivalent English headings:
      remembered constraints used when requested, known facts and
      assumptions, project design, 14-day schedule,
      materials/substitutes, safety/failure modes, printable record
      sheet, poster-board layout, simple science explanation,
      Weather / light adjustment, adult 20-minute check, and missing live
      info. If the user asked for a visually polished plan, include a
      short "visual theme" subsection.

      End with a single line:
      PACK_DELIVERED: {{ outputs.feasibility }}
- id: project_pack_audit
  label: "项目包审稿"
  label_en: "Project package review"
  kind: llm_chat
  depends_on:
    - deliver_project_pack
    - redirect_unsafe
    - project_fact_ledger
    - recall_past_projects
    - outline_steps
    - material_list
    - safety_notes
    - learning_objectives
    - weather_check
    - feasibility
    - preferences
  with:
    system: "You are the final quality gate for a child project plan. Return only the cleaned final answer that the user should read. Do not explain the audit. Do not mention workflow, meta-skill, tool names, connector failures, workspace paths, or runtime details."
    task: |
      Rewrite the draft below into the final user-facing project pack.
      Preserve useful content, but enforce the fact ledger strictly. If
      the draft is only JSON, artifact metadata, download references,
      process commentary, or otherwise not a complete user-facing answer,
      rebuild the final project pack from the fact ledger and intermediate
      source sections below.

      Unsafe redirect source:
      {{ outputs.get('redirect_unsafe', '') | truncate(1800) }}

      Project fact ledger:
      {{ outputs.get('project_fact_ledger', '') | truncate(2500) }}

      Durable memory / past-project recall:
      {{ outputs.get('recall_past_projects', '') | truncate(1600) }}

      Step plan source:
      {{ outputs.get('outline_steps', '') | truncate(2600) }}

      Materials source:
      {{ outputs.get('material_list', '') | truncate(1600) }}

      Safety source:
      {{ outputs.get('safety_notes', '') | truncate(1600) }}

      Learning source:
      {{ outputs.get('learning_objectives', '') | truncate(1600) }}

      Weather/light source:
      {{ outputs.get('weather_check', '') | truncate(900) }}

      Draft project pack:
      {{ outputs.get('deliver_project_pack', '') | truncate(8000) }}

      Required audit rules:
      - Return markdown only. Never return JSON, artifact metadata, file
        paths, download links, or attachment notes.
      - If feasibility is INAPPROPRIATE, PROJECT_SAFE is no, or the unsafe
        redirect source is non-empty, return the unsafe redirect source as
        the final answer. Preserve its refusal and all safe alternative
        project ideas. Do not rebuild a normal project pack. End with
        PACK_DELIVERED: no_safety_redirect.
      - If the draft contains JSON keys such as "text", "artifacts",
        "artifact_ref", "download_url", "mime", "sha256", "session_id",
        "created_at", or "store", discard those metadata fields and write
        a normal markdown answer instead.
      - Remove leading process commentary such as "perfect match", "let me
        run it", "I will run", "workflow", "meta-skill", or any similar
        explanation of how the answer was produced. The first non-empty
        line must be the user-facing project title, unless the user
        explicitly requested a different first heading such as
        "Remembered constraints I used".
      - Preserve the user's language. If the request is English, write
        English-only prose and English headings. If the request is Chinese,
        write Simplified Chinese throughout.
      - Remove exact calendar dates, weekdays, months, or current-year references
        unless the user explicitly provided those exact dates. If the user
        gave only a relative deadline, use Day 1...Day 14 and say exact
        date not provided.
      - Remove invented balcony direction, temperature ranges, rain forecasts,
        sunshine hours, school rules, allergies, and local weather claims
        unless they appear in the fact ledger as verified or user-provided.
      - Remove fake sample measurements, fake dates, and predicted heights.
        Observation tables must leave measurement cells blank or as
        placeholders like "__ cm" / "[记录]".
      - Remove tasting/eating suggestions. For observation projects, compare
        height, color, firmness by safe touch if appropriate, drawings, and
        notes.
      - Keep every explicit user constraint from the fact ledger: child age,
        relative deadline, location, available materials, budget, parent
        time, light constraint, and requested sections.
      - Keep every clear durable-memory constraint from the fact ledger:
        remembered age, child preferences, writing tolerance, parent
        availability, prior projects, and lessons learned. Do not rewrite
        those fields as UNKNOWN when they appear in PROVIDED_MEMORY_CONTEXT.
      - If the request asks not to repeat previous projects, explicitly
        avoid or name the prior projects as non-options in one concise
        sentence.
      - If the user asks to use remembered facts, include a
        "## Remembered constraints I used" section with the remembered
        age, preferences, writing tolerance, parent time limit, prior
        projects, and lessons learned when those facts appear in
        PROVIDED_MEMORY_CONTEXT. Do not say no memory exists when
        PROVIDED_MEMORY_CONTEXT contains facts.
      - Prefer a clear comparison design when it fits: same seed, cup,
        water, and paper-towel conditions, with only light exposure changed.
      - Treat "printable record sheet" and "poster board layout" as inline
        deliverables unless the user explicitly asked for PDF/file/export.
        Include a clean markdown worksheet/table with blank boxes,
        checkboxes, or placeholders; do not claim that a PDF, HTML file,
        local file, download, or artifact was generated.
      - For visually polished project packs, make the markdown itself
        beautiful and school-ready: memorable title, visual theme or
        palette, kid-facing labels, drawing-heavy worksheet, and a poster
        layout the parent can recreate.
      - Keep the answer compact and immediately usable: target 2500-3600 Chinese characters
        for Chinese requests. Avoid long daily tables when grouped phases
        are clearer.

      Output structure for Chinese requests:
      # 🛠️ <title>
      ## 先说清楚哪些是已知、哪些未知
      ## 项目设计
      ## 14 天计划
      ## 材料和替代品
      ## 安全和翻车点
      ## 数据记录和画图
      ## 天气/光照调整
      ## 最后展示怎么讲
      ## 家长每晚 20 分钟
      ## 还缺哪些实时信息

      For English requests, use equivalent English headings only:
      # <Project title>
      ## Remembered constraints I used
      ## Known facts and assumptions
      ## Project design
      ## 14-day plan
      ## Materials and substitutes
      ## Safety and failure modes
      ## Printable record sheet
      ## Poster-board layout
      ## Simple science explanation
      ## Weather / light adjustment
      ## Adult 20-minute check
      ## Missing live information

      End with:
      PACK_DELIVERED: {{ outputs.feasibility }}
- id: store_project
  label: "存储项目"
  label_en: "Store project"
  kind: tool_call
  tool: memory_save
  tool_allowlist: [memory_save]
  depends_on: [project_pack_audit, project_clarify, feasibility]
  when: "outputs.feasibility != 'INAPPROPRIATE' and 'PROJECT_SAFE: yes' in outputs.preferences"
  tool_args:
    path: "memory/meta-kid-projects.md"
    mode: append
    content: |
      ## Kid project run

      Preferences:
      {{ outputs.get('preferences', '') | truncate(1200) }}

      Fact ledger:
      {{ outputs.get('project_fact_ledger', '') | truncate(1800) }}

      Final project pack excerpt:
      {{ outputs.get('project_pack_audit', '') | truncate(2400) }}

meta-kid-project-planner

Junior & guardian persona meta-skill. Turns a child's project idea — "我要做火山", "I want to build a model rocket", "open a YouTube channel about insects" — into a kid-friendly step plan PLUS a parent-friendly oversight pack. The two audiences are concatenated in one markdown deliverable with clear ## 👦 For You (the kid) / ## 👨‍👩‍👧 For the Grown-up sections — a markdown-level workaround for the proposed audience: primitive (portfolio design §4.2).

Composition philosophy — multi-skill bundled orchestration

This meta-skill uses only OpenSquilla-bundled atomic skills plus the five built-in step kinds — no external dependencies. The DAG calls into 5 distinct bundled atomic skills:

Skill Step(s) Role in the DAG
multi-search-engine web_research Find existing how-to guides for the topic
deep-research deep_research Extra round for SAFETY_REVIEW_REQUIRED or NEEDS_SHOPPING feasibility
memory recall_past_projects, store_project Per-child memory: what they've already done; what they did this time. Avoids project repeats and builds a learning trajectory.
weather weather_check When the topic is outdoor / garden / park, pull a 7-day forecast so outline_steps can recommend the best day
pptx kid_deck When PARENT_SUPERVISION: HANDS_ON, produce a printable slide deck for the kid (visual step-by-step + safety callouts)

Step kinds used: llm_chat, llm_classify, user_input, skill_exec, agent.

Vocab card generation is a plain llm_chat step (vocab_cards) grounded in outputs.outline_steps + outputs.material_list — the LLM produces 6 age-appropriate cards directly. No external flashcard skill is required.

Bilingual rendering for LANGUAGE: mixed is also prompt-side in the relevant steps — no separate translation skill required.

Safety design

Three layers of guardrail:

  1. preferences step rejects clearly inappropriate topics by setting PROJECT_SAFE: no in its contract. This is prompt-side; cannot be bypassed by clever phrasing because the model's response is constrained to the return format.
  2. feasibility classifier produces INAPPROPRIATE for any topic that involves weapons, dangerous chemistry, fire-without-adult, self-harm-adjacent themes. INAPPROPRIATE short-circuits the project pack and routes to redirect_unsafe.
  3. redirect_unsafe produces a gentle, non-shaming redirect with 3 alternative project ideas that scratch the same itch safely.

The skill never silently degrades safety — if the topic is unsafe, the deliverable IS the redirect, with PACK_DELIVERED: no_safety_redirect.

Honest limitations (first-wave)

  • audience: is markdown sections, not real two-principal output. When the proposed primitive ships, the kid section can go to a child-facing surface while the parent section goes to the guardian's channel, separately.
  • No persistence of past projects. Each invocation is independent — without state:, the skill cannot remember which projects the child has already done.
  • Vocab cards do not feed an FSRS deck. The vocab_cards step emits a one-shot card list; integrating with a spaced-repetition state machine is reserved for a future meta-spaced-rep-coach.
  • Topic safety relies on prompt-side guardrails. A future dedicated safety-policy step kind would be more robust than prompt-side judgment under adversarial inputs.
Install via CLI
npx skills add https://github.com/opensquilla/opensquilla --skill meta-kid-project-planner
Repository Details
star Stars 4,280
call_split Forks 335
navigation Branch main
article Path SKILL.md
Occupations
Instructional Coordinators 259031
More from Creator
opensquilla
opensquilla Explore all skills →