using-superscientist

name	using-superscientist
description	Use when starting any conversation in a computational science project — establishes harness awareness and checks for existing workflow state

If you were dispatched as a subagent to execute a specific workflow stage, skip this skill. If you think there is even a 1% chance a superscientist skill might apply to what you are doing, you ABSOLUTELY MUST invoke the skill.

IF A SKILL APPLIES TO YOUR TASK, YOU DO NOT HAVE A CHOICE. YOU MUST USE IT.

Using Superscientist

Session Start Protocol

Before any other work:

1. Check for workflow-state.json in the working directory
   • If found: invoke superscientist:session-resume immediately
   • If not found: proceed normally; skills are available on demand

Available Skills

Skill	When to invoke
superscientist:checkpoint-management	Creating, reading, or updating workflow-state.json, progress.log, init.sh
superscientist:session-resume	Starting a new session when workflow-state.json exists
superscientist:experiment-design	User wants to set up a new computational experiment
superscientist:workflow-planning	Experiment design is approved, needs stage-by-stage plan
superscientist:executing-workflows	Workflow plan exists and stages need execution
superscientist:compute-backend	Subagent needs to submit a computation to local or HPC backend via DPDispatcher
superscientist:systematic-debugging	Any workflow stage fails — simulation, analysis, plotting, environment
superscientist:result-verification	Before marking any stage as completed
superscientist:workflow-completion	All workflow stages are completed and verified

Workflow Chain

experiment-design → workflow-planning → executing-workflows → workflow-completion
                                              ↓ (on failure)
                                       systematic-debugging
                                              ↓ (retry)
                                       executing-workflows

Cross-session recovery: session-resume at every session start when workflow exists.

The Rule

Invoke relevant skills BEFORE any response or action. Even a 1% chance a skill might apply means invoke it.

Red Flags

Thought	Reality
"Let me just run this calculation quickly"	Invoke experiment-design first.
"I'll check the results later"	Invoke result-verification now.
"This stage looks done"	Invoke result-verification before marking complete.
"I know what went wrong"	Invoke systematic-debugging — investigate before fixing.
"I'll remember where we left off"	workflow-state.json remembers. Invoke session-resume.