clauzz-cli

README

clauzz

Workspace context manager for AI coding agents.

AI coding agents scatter your work across sessions identified only by anonymous UUIDs. clauzz turns them into a managed workspace: give sessions memorable names, list them grouped by directory, search across all of them, resume any of them in one keypress, and carry context from one session into another.

Today clauzz supports Claude Code; adapters for other agents are on the roadmap.

Install

Linux / macOS:

curl -sSL https://clauzz.muzz-ai.com/install.sh | sh

The script downloads the latest release binary for your platform, verifies its checksum, installs it, and installs the Claude Code slash commands. Windows is not supported (resume uses exec(2)).

With Go installed:

go install github.com/ghulammuzz/clauzz-cli/cmd/clauzz@latest

Or build from source:

go build -o clauzz ./cmd/clauzz && mv clauzz /usr/local/bin/

Uninstall

curl -sSL https://clauzz.muzz-ai.com/uninstall.sh | sh

Removes the binary and the slash commands. The session registry at ~/.clauzz is kept; add | sh -s -- --purge to remove it too.

Slash command (optional)

To use /clauzz:add-session {name}, /clauzz:context {id-prefix}, and /clauzz:list from inside Claude Code:

mkdir -p ~/.claude/commands/clauzz && cp claude-command/*.md ~/.claude/commands/clauzz/

Usage

Command	What it does
clauzz	Interactive picker. Enter resumes the session via claude --resume in its directory
clauzz add {name}	Register the current Claude session under a custom name
clauzz list	Plain list of registered sessions, grouped by directory; ls is an alias
clauzz prune	Remove all [gone] entries (sessions whose transcript was deleted)
clauzz search {query}	Full-text search across every session on the machine, registered or not
clauzz rename {id-prefix} {new-name}	Rename a registered session
clauzz context {id-prefix}	Print a context digest of a session (used by /clauzz:context)
clauzz rm {id-prefix}	Remove a session from the registry (min 4 chars of the session ID); delete is an alias
clauzz --help	Show help

Example:

$ clauzz list
/Users/me/code/app
  Task Kafka                     625e4b2e   2026-07-09 10:12
  Task DB Replica                84409ceb   2026-07-08 21:30
/Users/me/code/app/membership
  Feat Membership List           bdd3bcef   2026-07-09 09:01

$ clauzz rm 8440
removed "Task DB Replica" (84409ceb) in /Users/me/code/app

How it works

• Registry lives at ~/.clauzz/sessions.json.
• add resolves the current session from $CLAUDE_SESSION_ID, falling back to the newest session transcript in ~/.claude/projects/{encoded-cwd}/.
• Entries whose transcript was deleted show [gone] and cannot be resumed; remove them with clauzz rm.
• Removing an entry never touches the Claude session itself.
• /clauzz:context {id-prefix} [focus query] injects a digest of another session into the active one: its title, every user prompt, and the last 20 messages (truncated). With a focus query (e.g. /clauzz:context 948e consumer group setup), Claude also greps the source transcript for that topic and loads only the relevant parts. Without one, it falls back to the transcript only when the digest is not enough.

Context transfer flow

How /clauzz:context moves context from session B into the active session A:

sequenceDiagram
    actor User
    participant A as Claude session A
    participant CLI as clauzz CLI
    participant Reg as ~/.clauzz/sessions.json
    participant B as Session B transcript (jsonl)

    User->>A: /clauzz:context {id-prefix-B}
    A->>CLI: clauzz context {id-prefix-B}
    CLI->>Reg: resolve prefix to session B entry
    CLI->>B: parse transcript
    Note over CLI,B: keep user prompts + assistant text,<br/>drop tool calls, results, thinking, sidechains
    CLI-->>A: digest (title, all user prompts,<br/>last 20 messages, transcript path)
    Note over A: digest becomes part of<br/>session A's context

    opt focus query given, or digest not enough
        A->>B: Read/Grep specific parts of the transcript
        B-->>A: only the details needed
    end

    A-->>User: summary of loaded context,<br/>ready to work with it