@telora/telora-ai

AI work tracking CLI for Telora. Registers workstations, logs AI interactions, and syncs work summaries to the Telora backend.

Requires Python 3.8+.

Installation

npm install -g @telora/telora-ai

Setup

1. Register your workstation

Create an AI tracker in Telora Settings > AI Trackers (one per workstation), then register with the code:

telora-ai register <registration-code>

2. Verify

telora-ai status

Commands

| Command | Description | |---------|-------------| | register <code> | Register workstation with Telora backend | | status | Show registration and tracking status | | track <issue_id> | Link work to a specific issue | | untrack | Stop tracking current issue and sync | | sync | Sync pending work summaries to Telora | | resync [issue_id] | Re-aggregate and re-sync previous work | | queue | Manage sync retry queue | | context | Show active issue context | | log <kind> | Log interaction events (used by hooks) | | activate <key> | Set the active issue (called by the hook) | | deactivate | Clear the active issue | | work-start | Log start of work on the active issue | | work-end | Log end of work with git artifacts | | config <key> <val> | Set configuration (e.g., telora_url) | | setup-hooks | Configure Claude Code hooks automatically |

Claude Code Hooks

This package includes hooks for automatic interaction logging:

• telora-ai-hook -- PostToolUse hook. Activates an issue when it moves to In Progress through the Telora MCP tool, deactivates it when it leaves In Progress, and logs tool calls while one is active
• telora-ai-human-hook -- UserPromptSubmit hook (logs human turns)

Configure in your Claude Code settings or .claude/settings.json:

{
  "hooks": {
    "PostToolUse": [{ "command": "telora-ai-hook" }],
    "UserPromptSubmit": [{ "command": "telora-ai-human-hook" }]
  }
}

With no active issue the hooks exit without spawning telora-ai, so they are cheap to leave on every tool call.