From antigravity-awesome-skills
Runs a structured clarifying interview for new project requests, then outputs a fully specified prompt.md for a fresh agent session to execute, preventing expensive mistakes from vague specs.
How this skill is triggered — by the user, by Claude, or both
Slash command
/antigravity-awesome-skills:brave-manThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
Most people describing a project ("vibe coders" included) only give a brief or partial picture of what they want. They can't be expected to specify everything up front — humans don't think in complete specs, and even when they try, they forget the small details that turn into real problems once the project has grown. If the agent starts building from a thin description, it fills the gaps with s...
Most people describing a project ("vibe coders" included) only give a brief or partial picture of what they want. They can't be expected to specify everything up front — humans don't think in complete specs, and even when they try, they forget the small details that turn into real problems once the project has grown. If the agent starts building from a thin description, it fills the gaps with silent guesses, and by the time those guesses turn out wrong, they're expensive to undo.
Brave Man flips the order: clarify exhaustively first, build later. The agent's job here is NOT to write code, scaffold files, or produce an implementation plan. Its only job is to run a structured interview until the project is fully understood, then write that understanding down as a single, clean, self-contained prompt.md file that a fresh agent session can execute later.
prompt.md. Do not generate an implementation plan artifact, scaffold a repo, or write any application code in this skill.prompt.md, and ask the agent to execute it.Never skip straight to building because the request "sounded simple." Simple requests still go through triage — triage is what decides how short the interview gets, not whether it happens.
Ask 2-3 quick questions before anything else, to calibrate depth:
Use the answers to decide which phases below need full depth, which need only one or two quick questions, and which can be skipped entirely with a stated default (e.g., a single static page skips Integrations & Auth entirely rather than asking about it).
Work through these one phase at a time. Within a phase, ask questions in one batched round (3-5 questions), not one at a time. Skip or shrink phases that triage marked irrelevant — say so explicitly ("skipping auth since this has no accounts") rather than silently dropping them.
(skip entirely if triage shows no accounts/external services needed — state that explicitly instead of asking)
Keep a running, visible status of each relevant phase using this format, and show it to the user as phases close:
[x] Purpose & users — confirmed
[x] Core features & flows — confirmed
[~] Data & content model — defaulted (assumed simple per-user storage, no sharing)
[ ] Tech stack & environment — open
[-] Integrations & auth — skipped (no accounts needed)
...
Do not move to synthesis while any relevant phase is still [ ] open. [x] confirmed and [~] defaulted-and-accepted both count as closed.
Once every relevant phase is closed, stop asking questions. Do not produce an implementation plan, do not scaffold a project, do not write application code. Instead, write a single file named prompt.md in the project root containing the full, distilled specification, addressed directly to whichever agent will read it next. Structure it as:
# Project Brief: <name>
You are building the following project. Treat this file as the complete
specification — everything needed to build it correctly is below.
Do not re-ask the questions that produced this brief unless something
here is genuinely ambiguous or missing.
## Overview
<one paragraph: what it is, who it's for, what success looks like>
## Core Features (prioritized)
<must-have list, then nice-to-have list>
## User Flows
<step-by-step walkthroughs from Phase 2>
## Data Model
<entities, relationships, persistence rules from Phase 3>
## Tech Stack & Environment
<language/framework, hosting/platform, repo constraints from Phase 4>
## Integrations & Auth
<or "None — no accounts or external services required">
## Non-Functional Requirements
<scale, sensitive data handling, hard constraints from Phase 6>
## Edge Cases & Error Handling
<from Phase 7>
## Assumptions & Defaults Used
<every default that was proposed and accepted during the interview,
listed plainly so the user can spot anything they want to override later>
## Definition of Done
<acceptance criteria and explicit out-of-scope items from Phase 8>
## Suggested Build Order
<a short, sensible milestone sequence — not a full implementation plan>
Keep it tight and complete rather than padded — every section should contain real decisions, not filler. The "Assumptions & Defaults Used" section matters most: it's the paper trail for every gap the user couldn't have specified up front.
After writing prompt.md, tell the user, plainly:
Your project spec is saved as
prompt.md. For the best results, start a new chat, tag this file, and tell the agent to execute it. Starting fresh keeps the build conversation free of the back-and-forth that produced the spec — the agent only needs the distilled brief, not the full interview, which keeps things faster and avoids burning context on a conversation it doesn't need anymore.
Do not start implementing in the current session even if the user asks immediately after — point them to the new-chat handoff, since that's the whole point of separating interview from execution.
1. **Triage:** Is this just for you? How big is it? Any preferred stack?
2. **Phase 1 (Purpose):** What is the one thing it absolutely must let you do?
3. **Synthesis:** Outputs `prompt.md` with React/Firebase stack based on interview.
Symptoms: User replies with "just build it" or "I don't care".
Solution: Stop asking questions, propose defaults for the remaining critical phases (Data, Auth), and synthesize the prompt.md.
@brainstorming - Use when exploring abstract ideas rather than gathering a build specification.prompt.md must be executed in a fresh agent session to ensure clean context.npx claudepluginhub sickn33/antigravity-awesome-skills --plugin antigravity-bundle-aas-python-api-builderClarifies vague requests via Socratic questioning, focusing on one key uncertainty (goals, scope, constraints, completion criteria) at a time to produce actionable requirements. Activates on deep-interview requests or unclear specs.
Orchestrates structured Socratic interviews to clarify ambiguous requirements using a dedicated interviewer agent and Ambiguity Score. Useful for vague ideas; invoke via /deep-interview or keywords.
Transforms vague goals into structured requirements via systematic interview. Three phases: Interview, Extract, Cross-check. Outputs requirements.md consumed by /blueprint.