iloom-cli

iloom-cli is a CLI tool built for Claude Code that creates isolated development environments (git worktrees, database branches, unique ports) for each task while documenting the AI agent's reasoning in your issue tracker. Instead of chat logs that disappear, iloom records analysis, execution plans, assumptions, and risks as permanent issue comments, so your team can understand why code was written a certain way and pick up context when context-switching between tasks. When tasks grow larger, swarm mode decomposes work into child issues and launches parallel agents to execute them independently, each with full iloom workflow visibility. Teams and solo developers benefit from persistent decision records, reduced cognitive overhead, and the ability to multitask across several agents without losing the thread.

1. **Install iloom-cli**: Run `pip install iloom-cli` or `brew install iloom-cli` depending on your system. Verify with `iloom --version`. 2. **Initialize Workspace**: Use `iloom init --name [PROJECT_NAME] --issue-tracker-path [PATH]` to create an isolated workspace. Replace [PROJECT_NAME] with your project (e.g., 'auth-migration') and [PATH] with your issue tracker directory (e.g., '~/projects/issue-tracker'). 3. **Generate Artifacts**: Run `iloom plan --requirements "[REQUIREMENTS]" --output-format markdown` to create a project plan. Replace [REQUIREMENTS] with your project needs (e.g., 'Migrate legacy auth to OAuth2 with PKCE'). 4. **Save to Issue Tracker**: Use `iloom save --file [PLAN_FILE.md] --issue-tracker-path [PATH]` to persist the plan. For GitHub, run `iloom github create-issue --title "[TITLE]" --body-file [PLAN_FILE.md]`. 5. **Reopen Workspace**: Later, use `iloom open --workspace [PROJECT_NAME] --issue-tracker-path [PATH]` to resume work. Tip: Use `iloom list` to see all active workspaces.