Mission belongs in the same broader movement as Spec Kit, BMAD, and GSD: make AI-assisted development less improvised and more reliable. The difference is the layer where Mission solves the problem.

Spec and prompt systems improve the agent’s instructions. Mission adds an operating layer around the agent.

The Short Version

Tool shape	Typical center	Mission difference
Prompt templates	better initial instructions	Mission owns runtime state and gates
Spec generators	better planning documents	Mission turns artifacts into executable workflow law
Agent chats	one interactive session	Mission splits work into bounded Agent executions
CLI task adapters	scripted automation	Mission keeps daemon state, worktrees, and command views

Mission does not compete by being a better model. It coordinates models, repositories, artifacts, and operators under one governed Mission flow.

What Mission Adds

Dimension	Prompt/spec tools	Mission
Workflow authority	The agent follows written instructions	The daemon enforces workflow state and legal commands
State	Chat history and markdown files	Repository control state, Mission runtime data, and Entity records
Execution safety	Often the active checkout	Isolated Mission worktrees
Recovery	Re-read docs and reconstruct context	Reconnect to persisted daemon-owned Mission state
Human control	Correct the agent through more chat	Pause, stop, relaunch, rework, verify, and deliver through operator commands

Why That Matters

Mission agrees with the core insight behind the other tools: raw improvised chat is not a reliable way to ship software. Where it diverges is in saying that the workflow itself should have an operating system.

That gives teams a durable control boundary around the whole Mission, not just better prompts inside one execution. The selected coding agent still does the work. Mission decides what work exists, when it can run, where it runs, what evidence is required, and how the operator stays in control.

Choose Mission when the real problem is not just better prompting, but controlling AI delivery as an observable, recoverable, operator-run system.