prompts.chat: From Prompt Collection to a Self-Hosted Library
Purpose, mechanisms, getting started and boundaries
Source: f/prompts.chat · f
User Problem
Positioning and task fitf/prompts.chat extends a prompt collection with browsing, contribution, versioning, self-hosting and integrations. Current documentation also includes discovery of bots and bot teams. It organizes prompts and related resources rather than making a saved template a reliable agent by itself.
- Manage recurring personal or team prompts
- Preserve versions and discussion around prompt changes
- Retrieve reusable resources through MCP or plugins
Product Mechanism
Capabilities and execution- Web and data formats coexistPrompts are available through a website, Markdown, CSV and dataset paths. Interactive browsing and programmatic processing can use related material, with field and update checks still required. Source
- Version prompts collaborativelySelf-hosting documentation describes version history, change requests and private prompts, supporting reviewable working assets rather than scattered copied snippets. Source
- Self-hosting has infrastructure needsSetup configures branding, theme and authentication, with PostgreSQL for storage. A private library needs database, authentication and backup management beyond project creation. Source
- MCP and bot discovery broaden accessMCP exposes prompts, skills, bots and teams. Directory entries describe installation and execution location, while actual agent capabilities belong to their underlying projects and environments. Source
Workflow
Start with a small task- Start with recurring templatesSelect recurring tasks and record input, output and limitations for each prompt before deciding whether a full platform is needed.
- Choose public use or private deploymentPublic browsing supports discovery; internal materials may justify self-hosting. Prepare the documented runtime, database and authentication first.
- Establish versions and categoriesInspect imported content, define purpose and explain revisions. Tags support retrieval while examples support quality evaluation.
- Connect an agent lastVerify manual retrieval and access before enabling MCP or plugins. Distinguish reading resources from writing library changes.
The trial sequence and checks are editorial guidance. Entry commands follow official documentation; this project was not executed for the article. Source
Official entry example
npx prompts.chat new my-prompt-library
cd my-prompt-libraryComplete the prerequisites above first. Examples contain no real credentials.
Limits and Boundaries
Constraints and selection| Boundary | Meaning | Suggested action |
|---|---|---|
| Self-hosting is not entirely local processing | Optional semantic search uses a model API and authentication can use external providers. Source | Document data flow for database, authentication, search and agents separately. |
| Popularity is not effectiveness | Voting and discovery help find material without proving fit for a model or task. Source | Evaluate real inputs and record compatible models and failure cases. |
Scroll horizontally to view the full table
It fits teams treating prompts as shared assets needing retrieval and collaboration. A small personal collection may only need a repository.
How to Evaluate Adoption
Trial scenario and verificationA team can collect requirement clarification, test drafts and release-note prompts, each with anonymized examples and expected outputs. Compare revisions on the same cases and record improvements and tradeoffs so version history captures knowledge rather than timestamps alone.
Verify retrieval, private access, version traceability and database recovery. For MCP, also check returned content and write scope.
Evidence: Pinned README · Supporting official documentation · Current repository。
Source fetched (UTC): 2026-11-02T07:29:44.049Z；snapshot commit: d2a57a827616。The source date denotes the repository snapshot commit, not the project’s original launch.