npm package discovery and stats viewer.

Discover Tips

  • General search

    [free text search, go nuts!]

  • Package details

    pkg:[package-name]

  • User packages

    @[username]

Sponsor

Optimize Toolset

I’ve always been into building performant and accessible sites, but lately I’ve been taking it extremely seriously. So much so that I’ve been building a tool to help me optimize and monitor the sites that I build to make sure that I’m making an attempt to offer the best experience to those who visit them. If you’re into performant, accessible and SEO friendly sites, you might like it too! You can check it out at Optimize Toolset.

About

Hi, 👋, I’m Ryan Hefner  and I built this site for me, and you! The goal of this site was to provide an easy way for me to check the stats on my npm packages, both for prioritizing issues and updates, and to give me a little kick in the pants to keep up on stuff.

As I was building it, I realized that I was actually using the tool to build the tool, and figured I might as well put this out there and hopefully others will find it to be a fast and useful way to search and browse npm packages as I have.

If you’re interested in other things I’m working on, follow me on Twitter or check out the open source projects I’ve been publishing on GitHub.

I am also working on a Twitter bot for this site to tweet the most popular, newest, random packages from npm. Please follow that account now and it will start sending out packages soon–ish.

Open Software & Tools

This site wouldn’t be possible without the immense generosity and tireless efforts from the people who make contributions to the world and share their work via open source initiatives. Thank you 🙏

© 2026 – Pkg Stats / Ryan Hefner

cc-dejavu

v0.4.1

Published

CLI tool to search and browse Claude Code bash command history

Readme

deja

CI codecov License: MIT Bun TypeScript

Search and browse your Claude Code bash command history.

"What was that docker command I ran yesterday?"
$ deja search docker --limit 4

[ok] docker build --no-cache --platform linux/amd64 -t ghcr.io/user/api-service:latest .
     Rebuild without cache for production
     12/30/2025, 12:46 AM | ~/projects/api-service

[ok] docker build -t api-service:test .
     Build test image
     12/30/2025, 12:45 AM | ~/projects/api-service

[ok] docker run --rm api-service:test npm test
     Run tests in container
     12/30/2025, 12:46 AM | ~/projects/api-service

[ok] docker push ghcr.io/user/api-service:latest
     Push to registry
     12/30/2025, 12:48 AM | ~/projects/api-service

Every bash command Claude Code runs is logged in session files. deja indexes them into a searchable database so you can find that command you ran last week, see what worked, and avoid repeating mistakes.

Features

  • Smart ranking - Combines BM25 relevance with frecency (frequency + recency) so the most useful commands appear first
  • Full-text search - Uses SQLite FTS5 for fast, typo-tolerant searching
  • Match highlighting - Search patterns are highlighted in yellow in the output
  • Automatic sync - History is automatically synced before each search

Install

curl -fsSL https://raw.githubusercontent.com/Michaelliv/cc-dejavu/main/install.sh | bash

Or with Bun:

bun add -g cc-dejavu

Manual download: Get binaries from releases

From source:

git clone https://github.com/Michaelliv/cc-dejavu
cd cc-dejavu
bun install && bun run build

Usage

# Search for commands containing "docker"
deja search docker

# Use regex patterns
deja search "git commit.*fix" --regex

# Filter by project directory
deja search npm --cwd /projects/myapp

# Filter by current directory
deja search npm --here

# Sort by time instead of frecency
deja search npm --sort time

# List recent commands
deja list
deja list --limit 50

# List commands from current project only
deja list --here

# Manually sync (usually automatic)
deja sync
deja sync --force  # Re-index everything

Examples

Find a failing build command:

$ deja search "npm run build" --limit 5
# Look for [error] entries to see what failed

What commands did I run in this project?

$ deja list --here --limit 20
# Shows only commands from current directory

What commands did I run in a specific project?

$ deja search "" --cwd /projects/api --limit 20
# Empty pattern matches everything, filtered by directory

Re-run something from last week:

$ deja list --limit 100
# Scroll through recent history, copy what you need

Commands

deja search <pattern>

Search command history by substring or regex.

| Flag | Description | |------|-------------| | --regex, -r | Treat pattern as regular expression | | --cwd <path> | Filter by working directory | | --here, -H | Filter by current directory | | --limit, -n <N> | Limit number of results | | --sort <mode> | Sort by: frecency (default), time | | --no-sync | Skip auto-sync before searching |

deja list

Show recent commands, sorted by frecency by default.

| Flag | Description | |------|-------------| | --limit, -n <N> | Number of commands (default: 20) | | --here, -H | Filter by current directory | | --sort <mode> | Sort by: frecency (default), time | | --no-sync | Skip auto-sync before listing |

deja sync

Index new commands from Claude Code sessions.

| Flag | Description | |------|-------------| | --force, -f | Re-index all sessions from scratch |

Ranking Algorithm

Results are ranked by combining BM25 relevance with frecency (frequency + recency):

score = frecency × (0.5 + normalized_bm25)
  • BM25: Search terms that are more prominent in a command score higher. "build" in bun run build ranks higher than "build" buried in a long command.
  • Frecency: Commands you run often and recently score higher. Uses logarithmic scaling for frequency and time-decay weights for recency.

This means a command that closely matches your search AND you use frequently will appear at the top.

Use --sort time to revert to simple timestamp ordering (ignores BM25 and frecency).

How It Works

Claude Code stores conversation data in ~/.claude/projects/. Each session is a JSONL file containing messages, tool calls, and results.

deja scans these files, extracts Bash tool invocations, and indexes them into a local SQLite database at ~/.cc-dejavu/history.db. It tracks file positions so subsequent syncs only process new content.

Auto-sync: By default, search and list automatically sync before returning results. Use --no-sync to skip this if you want faster queries.

Privacy: deja is read-only and local-only. It reads Claude's session files but never modifies them. No data is sent anywhere.

Data Model

Each indexed command includes:

| Field | Description | |-------|-------------| | command | The bash command that was executed | | description | What Claude said it does (e.g., "Build the project") | | cwd | Working directory when command ran | | timestamp | When the command was executed | | is_error | Whether the command failed | | stdout | Command output (stored, not displayed by default) | | stderr | Error output (stored, not displayed by default) | | session_id | Which Claude session ran this command |

For AI Agents

Run deja onboard to add a section to ~/.claude/CLAUDE.md so Claude knows how to search its own history:

deja onboard

This adds:

<deja>
Use `deja` to search bash commands from previous Claude Code sessions.

<commands>
- `deja search <pattern>` - Search by substring (or `--regex`)
- `deja list` - Recent commands
- `deja list --here` - Recent commands in current project
- `deja search <pattern> --here` - Search in current project
</commands>

<when-to-use>
- "Deploy like we did last time"
- "Run the same build command"
- "What was that curl/docker/git command?"
- "Set it up like we did on the other project"
- "Show me the failed builds"
- Looking up commands from previous sessions
</when-to-use>

<when-not-to-use>
- Finding files -> use Glob
- Searching file contents -> use Grep
- Commands from current session -> already in context
</when-not-to-use>
</deja>

Now Claude knows how to search its own history.

When to use deja

  • User asks "what was that command I/you ran?"
  • User wants to find a command from a previous session
  • User needs to recall commands from a specific project
  • User wants to see commands that failed

When NOT to use deja

  • Finding files by name -> use Glob
  • Searching file contents -> use Grep
  • Checking recent conversation context -> already in your context
  • User's personal shell history -> not indexed, only Claude's commands

Development

# Install dependencies
bun install

# Run directly
bun run src/index.ts search docker

# Run tests
bun test

# Run tests with coverage
bun test --coverage

# Build binary
bun run build

About the name

deja - short for deja vu, "already seen." It shows you commands you've already run.


MIT License