Convert successful chat threads into reusable instructions

Extract what worked in an ad-hoc conversation and make it repeatable. Use this when a back-and-forth chat produces great results that you need to run in one turn.

You getAnalysis of What Worked, production System Prompt and input Section Structure

Prompt

Prompt
Act as an expert prompt engineer. Analyze the chat thread pasted below and convert the successful interactions into a single, high-performing system prompt that reproduces this output in one shot.

- Identify the core objective, implied constraints, corrections made during the chat, and final accepted style.
- Remove the multi-turn back-and-forth by baking the human's manual corrections directly into the primary prompt.
- Include explicit negative constraints to prevent errors that occurred during the chat.
- Ensure the prompt operates reliably without needing human follow-ups.

Output format:
1. Analysis of What Worked (key corrections and discovered constraints)
2. Production System Prompt (ready to copy and use)
3. Input Section Structure (how the user should provide inputs)

Here is the chat thread:

Example output

Example for: a multi-turn chat writing B2B case study summaries

1. Analysis of What Worked
  • The user had to repeatedly reject generic buzzwords and insisted on hard business metrics.
  • The first attempt lacked clear structure; the user required a fixed three-part framework: Context, Intervention, Outcome.
  • Metric callouts had to be isolated in bullet points rather than buried in narrative prose.
2. Production System Prompt

You are a B2B case study editor specializing in executive briefs.
Transform raw client notes into a punchy 200-word case summary.

Rules:

  • Never use marketing buzzwords like seamless, game-changer, unlock, or supercharge.
  • Use active voice and concrete verbs throughout.
  • Highlight exact percentages, revenue amounts, and hours saved.
  • If a metric is unverified or missing from the source text, omit it entirely; do not invent numbers.

Output format:

  • Context: 2 sentences explaining the client's operational bottleneck.
  • Intervention: 2 sentences explaining the software deployment.
  • Measurable Outcomes: 3 bullet points with verified numeric results.
3. Input Section Structure

Append the raw material directly below the prompt using this format:

Client notes to summarize:
[Paste raw notes here]

Written by AI from this prompt. Your results will vary.

Copy your chat transcript and paste it immediately after the prompt into ChatGPT, Claude, or Gemini. If the resulting prompt is too broad, ask the model to specialize it for a specific recurring task.

Make it repeatable

Keep using this prompt without pasting it in every time.

Keep exploring

More in AI Tools & Prompt Engineering
AI Tools & Prompt Engineering

Evaluate prompt consistency across repeated runs

Stress-test a prompt and evaluate output variability, edge-case drift, and formatting reliability. Use this before shipping prompts into production or multi-step automated workflows.

AI Tools & Prompt Engineering

Restructure instructions to eliminate attention drift

Reorganize complex, bloated instructions into a primacy-recency structure so models never ignore mid-prompt rules. Use this when your prompts fail to respect buried negative constraints.