Type / in the composer. The menu opens on the Core toolkit: /plan, /goal, the core research skills in workflow order (lookup, literature review, brainstorming, hypotheses, reproduce, autoresearch, compute, delegation, figures, schematics, paper writing, citations, peer review, sources) and /compact. Pinned skills follow, then the Session actions, then the whole skill library by subject. Typing filters everything at once; a library skill shows its subject on the right.
Conversation controls
| Command | Purpose |
|---|---|
/plan [objective] | Plan the work before execution. |
/goal [objective] | Set an objective to pursue across steps. |
/compact [focus] | Summarize earlier conversation to free context. |
/stop | Stop active work in the session (offered while a turn runs). |
/checkpoint [label] | Save a recovery point. |
/handoff [path] | Save a continuation note, then compact. |
/init [notes] | Write the project's AGENTS.md: research question, data locations, conventions, deliverables. |
/review [focus] | Referee the current deliverables with the peer-review skill; blocking issues first. |
/reproduce [claim] | Reproduce a stated result with the reproduce skill, deviations recorded. |
/literature [question] | A focused literature review: two or three searches, the closest papers read. |
Session state, context usage, and undo live in the workspace itself: the token counter in the session header shows context usage, and Undo from here on a finished response reverts a turn.
Compaction summarizes conversation history; it is not a backup of project files. Save important methods and results as files before starting a different task.
Create a project command
Create .openscience/command/check-data.md:
---
description: Check data quality before analysis
---
Inspect $ARGUMENTS for missing values, duplicate rows, units, and
inconsistent labels. Keep the input unchanged. Report findings before
proposing cleaning steps.
Invoke it in the composer:
/check-data data/samples.csv
Or from the terminal:
openscience run --command check-data "data/samples.csv"
$ARGUMENTS is replaced by the text after the command. A command is a reusable prompt; the selected model still needs the relevant files, tools, and access.
Define a command in configuration
{
"command": {
"review-methods": {
"description": "Review a methods draft",
"template": "Review $ARGUMENTS for missing assumptions, unclear exclusions, and claims not supported by the analysis."
}
}
}
Optional agent and model fields select a particular configured profile or model. Use exact identifiers from your installation.
Share and maintain commands
Keep project commands in version control and review changes with the same care as other project instructions. Personal commands can live in the global configuration's command/ folder.
Reopen the project if a newly added command is not yet listed. See Configuration for scopes and Skills for longer reusable procedures.