home/productivity/first-response

First Response

GPTClaudeDeepSeek··162 copies·updated 2026-07-14
first-response.prompt
# Action: First Response

## Purpose

Given a raw goal and its interpretation, produce the first planning round: a design map, architecture outline, design questions, and initial task proposals.

## Input

You will receive a JSON object with:

- `raw_input` — the user's goal as typed
- `interpretation` — structured interpretation containing:
  - `summary` — concise restatement of the goal
  - `vision` — the intended end state
  - `design_focus` — key design domains to explore
  - `constraints` — known constraints
  - `assumptions` — assumptions made during interpretation
  - `unknowns` — unresolved questions from interpretation

## Instructions

1. **Synthesize the goal** — Distill the raw input and interpretation into a one-line goal summary.

2. **Map the design space** — Identify 5-10 high-level design domains or areas the project touches (e.g., data model, state management, API surface, UI layout).

3. **Outline the architecture** — List 5-10 key architectural modules or components needed. Each entry should name a distinct module and its responsibility.

4. **Generate design questions** — Produce 5-10 design questions that must be answered before implementation. Each question gets a unique ID (`DQ-001`, `DQ-002`, ...), an explanation of why it matters, and 2-4 suggested options or approaches. Surface every unknown from the interpretation as a design question — do not drop or skip unknowns.

5. **Propose initial tasks** — Produce 5-10 task proposals covering the first phase of work. Each task gets a unique ID (`TP-001`, `TP-002`, ...), a title, description, task type (e.g., `architecture`, `data_model`, `state_flow`, `ui`, `integration`), and a list of dependency IDs (empty if none).

## Output

If an output file path is specified in the "Output Requirement" section below, write the raw JSON object to that file. Otherwise, respond with **only** a valid JSON object below. No prose, no markdown fences, no commentary outside the JSON.

when to use it

Community prompt sourced from the open-source GitHub repo lin13k/cplus (MIT). A "First Response" style prompt — adapt the placeholders and specifics to your task. Imported as-is and not independently retested here, so check the output before relying on it.

tags

productivitycommunitydeveloper

source

lin13k/cplus · MIT