Automation

Debug iOS self-runs an investigation loop for iPhone and simulator bugs. It verifies the device and log daemon, drives the app with taps or gestures, and reports root cause findings for crashes, hangs, and wrong behavior.

Debug Ios runs an end-to-end investigation of iOS app bugs by checking device or simulator state, capturing logs, and driving the app with synthetic taps and gestures. Use it to reproduce crashes, hangs, spinners, or other wrong behavior and report likely root cause.

Kanban-Based Development coordinates autonomous work on a shared kanban-md board with task claims, handoffs, and merge-safe parallel execution. Use it when multiple agents or humans need to work through tasks on the same codebase.

Kanban-Based Development coordinates autonomous work on a shared kanban-md board. Use it to claim tasks, avoid duplicate work, and hand off blocked items for human input or review.

Kanban Md manages project tasks in Markdown files with a command-line kanban board. Use it to create, move, list, and track work items, priorities, blockers, and sprint status.

Kanban Md manages project tasks in Markdown files with a CLI. Use it to create, list, move, block, and track work items on a kanban board.

Step By Step turns tasks that need the user's hands into clear, one-action instructions they can follow without guessing. Use it when you need someone to run commands, click, approve, or provide a secret, and verify each step after they act.

Test Agent Hooks exercises a real provider hook turn and checks that Zedra moves through Running, WaitingApproval, and Completed states. Use it when validating Codex, OpenCode, or Pi hook integrations and Delta notifications.

Codebase Explorer scans a codebase in phases to identify architecture, components, layers, and organization, then summarizes findings with file:line evidence. Use it to understand unfamiliar systems before planning changes.

Dispatching Workflows executes a phased plan as reviewed multi-agent workflows, with TDD, mandatory in-harness review, and contrarian verification before each phase returns. Use it when each wave should be validated inside the harness rather than by a lone subagent.

Execute is used when a task is implementation-ready and needs code changes now. It guides the agent to confirm scope, edit the smallest necessary files, run checks, and report validation and follow-ups.

Executing Plans loads a plan, checks it critically, and runs tasks in autonomous or batch mode with code review checkpoints. Use it when implementation needs pauses for human feedback between batches.