tryosschat/tembo-cli

Overview

This skill provides a command-line interface to the Tembo Public API for creating and managing AI coding tasks, listing repositories, and integrating Tembo into automation workflows. It is built in TypeScript and is intended for developers who want to script Tembo interactions, embed Tembo into CI/CD pipelines, or build custom tooling around Tembo agents. The CLI wraps Tembo endpoints and handles authentication, task lifecycle, and repository management.

How this skill works

The CLI authenticates using an API key and issues HTTP requests to Tembo's public API (https://api.tembo.io). It supports read-only commands for listing tasks and repositories and state-changing commands to create tasks or manage credentials. Before any command that modifies state (like creating a task or logging in/out), the tool requires explicit user confirmation to prevent accidental operations and compute costs.

When to use it

• Automating creation of coding tasks for AI agents from scripts or CI jobs
• Listing, searching, and auditing recent Tembo tasks and activity
• Connecting and managing repository access for Tembo-powered workflows
• Triggering agent runs from GitHub Actions, GitLab CI, or other pipelines
• Building developer tooling that needs programmatic access to Tembo

Best practices

• Always confirm with the user before running commands that modify state (tasks create, auth login/logout).
• Store the API key securely using environment variables (TEMBO_API_KEY) or protected CI secrets.
• Use --json output in automation to parse results reliably and avoid fragile CLI parsing.
• Limit task creation frequency to avoid hitting rate limits (100 req/min, 1000 req/hr).
• Pin agents explicitly (for example claudeCode:claude-4-5-sonnet) to get consistent behavior across runs.

Example use cases

• In CI: create a Tembo task to fix failing tests for a pull request and attach the repo/branch.
• Daily audit script: list recent tasks and export JSON for reporting or billing reconciliation.
• Developer tooling: search past tasks for similar bug fixes before assigning a new task.
• Ops automation: programmatically enable or list repositories available to Tembo agents.
• Interactive use: create a focused bug-fix task with a specific agent and branch from the terminal.