Working with AI Conversations
Summarize a Chat for Handoff
Asks the assistant to write a handoff brief with goal, decisions, constraints and next steps, ready to paste into a new chat.
- Customizable fields
- 1 field
Working with AI Conversations
Writes a searchable title, tags, key terms and a one-line description for a finished chat, so it is easy to find months later.
Designed for general-purpose AI assistantsPublished November 2026
Fill in the fields and the prompt updates. What you type stays in this page and is never saved or sent.
Help me label this conversation so I can find it again in several months, when I will remember the problem I had but not the words used in the chat. How I organize my chats: none yet Words I would probably type when searching for this later: not sure Project or area this belongs to: Read the whole conversation and produce: 1. Title: at most 8 words, starting with the topic and then the type of result (for example "Postgres connection pooling: decision and settings"). It must make sense out of context. 2. One-line description of the outcome, written for someone who has never seen the chat. 3. Result type: choose one of Decision, Draft, Explanation, Code, Plan, Research, Troubleshooting, Brainstorm, or Other. 4. Status: Finished, Needs follow-up, or Abandoned, with one line on what is left if it is not finished. 5. Tags: up to 6 lowercase tags. Prefer the folders and tags I already use when they fit, and mark any new tag as [new]. 6. Search terms: ten words or phrases I might type later. Include the names of tools, products, error messages, people and projects that appear in the chat, and also the plain-language words for the problem that never appeared in it. 7. Follow-ups: dates, deadlines or promised actions mentioned in the chat, quoted exactly. Rules: use only what is in the conversation. If you cannot tell the outcome, say "outcome unclear" in the description rather than guessing. Do not include private details such as passwords, keys or personal data in any label.
Copy customized prompt includes your values. Copy original keeps the {{placeholders}}.
Illustrative only. Real answers depend on your input and the assistant you use.
Title: Postgres connection pooling: settings decision Description: Chose PgBouncer in transaction mode with a pool of 20 for the API; the checkout service keeps direct connections. Result type: Decision Status: Needs follow-up. Load test the pool size before release. Tags: postgres, pooling, pgbouncer, api, performance, [new] infra-decisions Search terms: too many connections, connection pool, pgbouncer transaction mode, max_connections, FATAL remaining connection slots, database timeouts under load, pool size, prepared statements pgbouncer, api scaling, checkout service Follow-ups: "Run the load test before the March release."
The plain-language search terms are the most valuable part. Chats are usually titled after what the assistant said, and you will search with the words you used for the problem.
If your tool only has a title field, put the topic and result type in the title and the search terms in the first line of the chat or in a note.
Do this when the work is fresh. Labeling a chat a month later means reading the whole thing again.
If you use HistChat, folders, tags and full-text search work across ChatGPT, Claude, Gemini and Grok conversations, so the tags and search terms from this prompt are useful in more than one assistant.
Working with AI Conversations
Asks the assistant to write a handoff brief with goal, decisions, constraints and next steps, ready to paste into a new chat.
Working with AI Conversations
Studies a conversation that went well and writes a reusable prompt template with variables, so you can repeat the result next time.
Working with AI Conversations
Converts a troubleshooting or learning chat into a clean step-by-step guide that leaves out the dead ends and keeps what worked.