prompt-vault

Solid

Save, search, tag, and manage reusable AI prompts via the `pv` CLI. Local SQLite-backed prompt library with version history, tag filtering, and JSON export/import.

AI & Automation 36 stars 22 forks Updated today MIT

Install

View on GitHub

Quality Score: 80/100

Stars 20%
52
Recency 20%
100
Frontmatter 20%
70
Documentation 15%
78
Issue Health 10%
50
License 10%
100
Description 5%
100

Skill Content

# Prompt Vault Local-first prompt manager. Use when asked to save a prompt, find a saved prompt, list prompts by tag, or export/import the prompt library. ## Prerequisites - `pv` CLI installed: `npm install -g BunsDev/prompt-vault` - Data stored at `~/.prompt-vault.db` (SQLite) ## Commands ### Save a prompt ```bash pv add "prompt text here" --tags tag1,tag2 ``` - Tags are comma-separated, no spaces - Returns the prompt ID ### Search prompts ```bash pv search "keyword" ``` - Substring match on prompt text ### List prompts ```bash pv list # all prompts, newest first pv list --tag code # filter by tag ``` ### Export library ```bash pv export # JSON to stdout pv export > prompts.json ``` ## Workflow Patterns ### Save a prompt the user liked When the user says "save that prompt" or "remember this prompt": 1. Identify the prompt text from conversation context 2. Ask for tags if not obvious (or infer from context) 3. `pv add "<text>" --tags <tags>` ### Find a prompt for reuse When the user says "find my prompt about X" or "use that refactor prompt": 1. `pv search "X"` to find matches 2. Present the best match 3. Use it in the current task ### Suggest tags Common useful tags: `system-prompt`, `code-review`, `refactor`, `summary`, `debug`, `explain`, `test`, `docs`, `creative`, `agent`, `tool-use` ## Notes - Prompt text with quotes: escape inner quotes or use single quotes around the CLI arg - The DB is local-only — no sync. Use `pv export` to b...

Details

Author
OpenCoven
Repository
OpenCoven/coven
Created
3 months ago
Last Updated
today
Language
Rust
License
MIT

Integrates with

Similar Skills

Semantically similar based on skill content — not just same category