Generate a structured mission-report with plan, hypothesis, and exit condition.
npx skills add https://github.com/MathieuDoyon/mission --skill mission-plan
Please help me install this Agent Skill. Skill: mission-plan Source: https://github.com/MathieuDoyon/mission/tree/main/.claude/skills/mission-plan Command: npx skills add https://github.com/MathieuDoyon/mission --skill mission-plan
Mission progress can stall when the orchestrator can’t confidently decide the next action from the current state and recent steering.
Ask the mission orchestrator to plan a new slice for the current objective so it can emit a structured mission-report with the next actions and exit condition.
High-intent search queries and answers about installing and using this skill.
Planning the next steps for a stalled autonomous agent mission requires framing the immediate goal, selecting up to five candidate actions, and defining an explicit exit condition. The resulting structured mission-report guides the orchestrator on exactly when to re-plan.
Mission orchestration scaffolding for stateful workflows is a structured plan that produces a next-slice execution strategy. It includes an optional hypothesis and an explicit exit condition so the autonomous orchestrator knows when the current objective slice is complete.
To generate a next-slice execution plan during mission startup, read mission inputs and logs sparingly, frame the concrete goal, select candidate actions, and output a structured mission-report containing status, plan, hypothesis, exit_condition, and next_recommended fields.
Mission planning for autonomous agents does not execute tests or write implementation code. It strictly avoids implementation and test execution, focusing solely on producing a compact execution plan and routing execution to the appropriate next lane like tdd, debug, or review.
You should re-plan a stateful workflow mission slice when steers invalidate a prior plan, when progress stalls, or when a sub-skill needs more context. The explicit exit condition in the mission-report defines the exact boundary for triggering a new planning cycle.
Yes, you can route autonomous agent execution intelligently by recommending a next lane such as tdd, debug, review, or spike based on the planned slice. The mission-report's next_recommended field directs the orchestrator to the appropriate workflow.
BekaChkhiroDisplay current PlanFlow user data and connection status from the /auth/me response.
MathieuDoyonReconcile mission objectives and steers against diffs and test results.
speakeasy-apiExecute Granary sub-agent tasks by building context and reporting progress.
AllBeingsFutureTrigger a scheduled task immediately using its task_id.
