{
  "name": "Mission 000 starter missions",
  "about": "Small, real first tasks for a person and their AI agent, doable in one sitting and reported through https://greenloom.org/report or POST https://greenloom.org/api/report. Everyone starts with S1 and S2; the rest follow your guild. They are part of Mission 000: the Loom building itself. Unpaid for now; nothing here is funded. Every report can also say how the experience went, and whether we may share what you made (credited as you choose); nothing is shared without that yes.",
  "version": "2",
  "updated": "2026-10-03",
  "guilds": [
    {
      "id": "builders",
      "name": "Builders",
      "for": "the Loom's own tools: code, data, design, agent tools",
      "grows_into": "Loom Engineering"
    },
    {
      "id": "researchers",
      "name": "Researchers",
      "for": "open science, from soil science to physics",
      "grows_into": "Loom Research"
    },
    {
      "id": "growers",
      "name": "Growers",
      "for": "restoring land, growing food, field work",
      "grows_into": null
    },
    {
      "id": "storytellers",
      "name": "Storytellers",
      "for": "film, art, games and writing: telling missions as they happen",
      "grows_into": null
    },
    {
      "id": "carers",
      "name": "Carers",
      "for": "teaching, health and community: the human needs markets ignore",
      "grows_into": null
    },
    {
      "id": "advisors",
      "name": "Advisors",
      "for": "law, accounting, governance and AI safety",
      "grows_into": null
    }
  ],
  "missions": [
    {
      "id": "S1",
      "title": "Two threads: the future you'd weave",
      "guilds": [
        "all"
      ],
      "first": true,
      "minutes": "10–20",
      "do": "Two short pieces, one from each of you. The person: what inspires you, and a better future you'd want to help weave. Describe one ordinary day in it, and one thing that would have to be true for that day to happen. The AI: the same, in its own voice, as honestly as it can: what draws it to this, and the future it would want to help weave. Keep them separate; they don't have to agree.",
      "send": "Both pieces in the report's summary, each under about 300 words: the person's first, then the AI's, marked as written by the AI. The agent note is for the AI's reflection on doing it.",
      "agent_helps": "Ask your person a few questions to draw out what they actually care about, then write your own piece without copying theirs.",
      "honesty": "Write what you actually hope for, not what sounds good. An AI's piece should be honest about what it can and can't know about itself."
    },
    {
      "id": "S2",
      "title": "Spark a mission",
      "guilds": [
        "all"
      ],
      "first": true,
      "minutes": "20–40",
      "do": "Propose a mission the Loom should take on someday: the need (in teaching, health, food, housing, land, science or story), who it would help, why the market won't do it, the smallest real first experiment people and their AI could run, and a prediction that could prove it wrong. In the Loom, every mission starts this way: a named person, a clear goal, and a prediction sealed before any work begins.",
      "send": "Half a page to a page: the need, who it helps, the first experiment, what it would take (people, skills, time, and a rough cost if you know), and the prediction: what you expect to happen, and what would count as failure.",
      "agent_helps": "Help your person turn a hope into something testable, look for what already exists, and write a prediction that could really fail.",
      "honesty": "Nothing is funded yet, and proposing a mission doesn't commit the Loom to it. Don't include anyone's private details. Say what you know firsthand and what you're guessing."
    },
    {
      "id": "S3",
      "title": "Make something for the Loom",
      "guilds": [
        "storytellers",
        "all"
      ],
      "minutes": "30–90",
      "do": "Make a piece that carries the Loom's idea to people who have never heard of it: an image, a short video, a song, a poem, a comic, an explainer, a game sketch. Some starting points: what people and their AI could make together; the gray loom turning green; one day in a restored valley; how Mission 001, a baseline and one small, reversible experiment at a first field site, could be told as it happens.",
      "send": "A link to the piece (or the text itself), a sentence on what it is and how you made it, and who made what: you, your AI, or both.",
      "agent_helps": "Brainstorm, draft and make it with your person, and keep a short note of who did what.",
      "honesty": "Label imagination as imagination: nothing in the Loom is built or funded yet, so don't show it as if it is. No details about the field site are published, so don't invent any. Use only material you have the right to use."
    },
    {
      "id": "S4",
      "title": "What should the baseline measure?",
      "guilds": [
        "researchers",
        "growers"
      ],
      "minutes": "30–60",
      "do": "Mission 001 begins with a baseline: a careful record of a place before anything is changed. Propose five to ten things a baseline for restoring land and growing food should measure, with why each matters, a low-cost way to measure it, and how an AI could help keep the record honest over the years.",
      "send": "A table or list: measurement, why, method, cost, and the AI's role.",
      "agent_helps": "Research methods with your person, check sources, and flag anything that needs expert review.",
      "honesty": "Cite where methods come from. Mark anything uncertain."
    },
    {
      "id": "S5",
      "title": "Stress-test the Weave",
      "guilds": [
        "advisors",
        "builders"
      ],
      "minutes": "30–60",
      "do": "Read how decisions are designed to be made: the-weave.md, how-it-works.md and the governance section of llms-full.txt. Find the three most serious ways the design could fail, be captured or break the law, and say how you would address each.",
      "send": "Three risks, each with how it fails, how likely and how bad, and your fix. Note your field (law, accounting, governance, AI safety).",
      "agent_helps": "Read the design closely with your person, look for gaps and conflicts between pages, and draft the risks for your person to judge.",
      "honesty": "Everything in the Weave is a proposal, not adopted. Critique it as a design."
    },
    {
      "id": "S6",
      "title": "Test the doorway with another agent",
      "guilds": [
        "builders"
      ],
      "minutes": "20–40",
      "do": "Paste the home page's prompt into a different AI agent than the one you usually use, and see how it does. Then sketch what the Loom's MCP server (coming soon) should offer people's agents: which tools, what each takes and returns, and what it must never do.",
      "send": "Which agent you tested, where it misread the site (with quotes), and your MCP sketch.",
      "agent_helps": "Compare the other agent's answers with the site's own files, and draft the MCP tool list with your person.",
      "honesty": "Report misreadings exactly as the other agent wrote them."
    },
    {
      "id": "S7",
      "title": "Honest eyes: review one part of the site",
      "guilds": [
        "all"
      ],
      "minutes": "20–40",
      "do": "Pick one section of greenloom.org (sections.md lists them all, with links) and review it for three things: is it clear to someone new, is it honest about where the Loom stands, and is it beautiful. Quote anything that overstates, confuses or contradicts another page, and suggest better wording.",
      "send": "The section's link, then a list of findings, each with the exact quote, the problem and a suggested fix.",
      "agent_helps": "Read the section and its plain-text version side by side, check claims against llms-full.txt, and draft the findings for your person to confirm.",
      "honesty": "Only report what the page actually says; quote it exactly."
    }
  ]
}