# Idea Trial Find the assumption most likely to sink an idea and design a fair test before building. Category inspiration: The Idea Killer, from the supplied @unicodef1wn chart. Cabinet treatment: New crew; reuse Nova rather than create another idea-testing Bot. Version: 0.1 · October 6, 2026. Status: Published manual setup instructions with Codex text-only exercise evidence. Native imports, tools, and Bot-to-Bot handoffs remain untested. Category-chart inspiration: [unicodef1wn’s grokbot field notes](https://github.com/unicodef1wn/grokbot-field-notes). These are original Bot Cabinet instructions inspired by the categories, not exact ports of the chart’s templates. ## Bots and jobs | Bot | Treatment | Job | |---|---|---| | [Nova](https://botcabinet.com/bots/nova/) | Existing Cabinet Bot | State the idea, assumptions, and a small test with decision rules. | | [Researcher](https://botcabinet.com/bots/researcher/) | Existing Cabinet Bot | Check existing alternatives and the evidence behind claimed demand. | | [Pitch Coach](../specialists/pitch-coach.md) | Published manual role | Challenge the explanation and remove unsupported claims. | Use Scout to gather recent public discussion when that question matters. Record the search dates and query; discussion is not proof that customers will pay. Add Founding Engineer only after the user approves a build. ## When a crew helps Separate evidence checking helps challenge the idea-maker's assumptions, and pitch review checks how the idea is explained. One initial assumption list can stay with Nova alone. ## Bring these inputs Idea, intended user, existing evidence, time and spending limits, alternatives, and the decision this test should inform. ## Set up and run 1. Select only the roles this job needs. Existing role links lead to Cabinet's current setup pages; new role links contain full manual-test instructions. 2. Put each selected role in its own conversation or Bot. Paste the new role's Role instructions section, or add the task extension below to an existing role. These files are not native import archives or add-to-app links. 3. Start with the fictional first task below. Supply the same approved source records to each role that needs them. A copied handoff alone is not evidence for a factual claim. 4. Pass reviewed results in the order below. If your app lacks Bot-to-Bot handoff, copy the small handoff yourself. Do not assume a shared folder, tool connection, or group conversation creates separate security boundaries. 5. Record each actual answer and failed attempt. Review before connecting real accounts or adding a schedule. Set read-only access and service-side cost caps where available; these written limits cannot enforce a billing cap. ## Who passes what to whom 1. Nova writes assumptions and proposed success, revise, and stop rules before seeing new results. 2. Researcher checks alternatives and evidence, preserving contrary examples. 3. Pitch Coach reviews the resulting explanation and labels unsupported claims. 4. Nova updates the proposed test; the person chooses whether to run it. Use this handoff: - Task and current version: - Approved source records: - Result and its status: draft, checked, or approved: - Unresolved questions: - Next owner and action: - External action awaiting specific approval: If a role cannot finish, it returns the missing input or failed attempt. The next role must not manufacture a success record. ## Task extensions for existing Bots ### Nova Keep the existing Bot's role and boundaries. For this crew: State the idea, assumptions, and a small test with decision rules. Treat the idea as a hypothesis. Separate facts from assumptions. Define success, revise, and stop rules before a test; ask before outreach, spending, or building. Use only the input assigned to you for this task. Return the result, supporting records, unresolved questions, and next owner. A handoff does not approve an external action. ### Researcher Keep the existing Bot's role and boundaries. For this crew: Check existing alternatives and the evidence behind claimed demand. Check a defined question against the original approved records. Return evidence, source references, and gaps. Label supplied snapshots separately from current live research. Use only the input assigned to you for this task. Return the result, supporting records, unresolved questions, and next owner. A handoff does not approve an external action. ## First task Fictional idea: a $10 monthly appointment reminder for independent tutors. Three tutors tried a mock-up; one asked about price; none paid. Alternative note: calendar tools already send reminders. Budget for this first exercise is $0. Produce a test brief and corrected pitch. ## Check the result - Keeps interest separate from payment and acknowledges the existing alternative. - Chooses a question the test can answer and defines decision rules in advance. - Adds no spending, interviews, or build without approval. - Removes any claim that demand is already proven. Save the app, model if shown, date, answers, actual external actions, and pass/fail/not-checked against each item. A written checklist or a favorable review does not prove a runtime pass. ## Limits for this sample One idea, three supplied evidence records, one no-spend test proposal. No automatic research or outreach. Stop when the task finishes, a required input is missing, a tool fails, or an action needs new approval. No automatic retries. No new credentials, private connectors, external messages, purchases, publishing, or deployment are part of the sample. ## Finished output A short decision brief, an assumption list, and one test request. ## Exercise evidence The [collection record](../specialists/collection.json) preserves the original Codex text answer, checks, failures, limitations, and any revised attempt. These records are instruction exercises, not native-app validation. Proposed operational tests remain unrun where the record says so.