terminal

Chat Controls & Commands

Master powerful chat commands to streamline your collaboration with Nova and 800+ AI experts. Use 60+ slash commands, agent references (@), and document controls (#) to work efficiently.

Meet Nova 🟣

Nova is your intelligent guide and coordinator chatbot—"a new star" that helps you navigate features, discover experts, execute workflows, and manage projects. Inspired by Claude Code and Vertex AI Studio, combining natural conversation with command-line efficiency.

code

Natural + Structured

Chat naturally or use slash commands for precision. The interface adapts to your workflow.

auto_awesome

Smart Autocomplete

Type / for commands, @ for agents, # for documents—autocomplete guides you.

memory

Context Aware

The system tracks your project, documents, and active tasks—commands work with your current context.

Command Syntax

Three simple prefixes unlock powerful workflow automation

/

Slash Commands

Invoke actions, verbs, or workflow tasks

/scatter "self-healing coating" --n=12
/evaluate Idea-24 --metrics novelty,feasibility
/draft #Spec_v2 --style=uspto-plain

Pattern: /verb arguments

@

Agent References

Tag agents, people, or roles to direct actions

@examiner please critique this
/assign draftClaims @claimsSpecialist
/add @examiner

Pattern: @agentName or @role

#

Document References

Reference documents or artifacts in workspace

/read #Spec_v2
/proofread #AbstractDraft
/merge #Idea12 #Idea17

Pattern: #docName or #artifactId

auto_fix_high Combined Usage

Combine slash, @ and # references in a single command for powerful precision:

/assign /proofread #Spec_v2 @proofreader
/do review #ClaimSetB --reviewers @examiner @advisor
/merge #Idea12 #Idea17 --by @strategist

Available Commands

60+ essential commands organized into 7 intuitive categories

/create [type]

Generate new idea, document, or task

/create idea "biodegradable packaging"

/read #document

Load and display a document

/read #Spec_v2

/update #document

Modify existing content

/update #ClaimSetA --section claims

/delete [item]

Remove item from workspace

/delete Idea-42

/list [type]

Show ideas, tasks, documents, or agents

/list ideas --tag=novel

/analyze #document

Perform deep analysis with optional focus

/analyze #Spec_v2 --focus novelty

/explain [concept]

Clarify a concept, decision, or agent action

/explain @examiner "why reject claim 3?"

/compare #doc1 #doc2

Side-by-side comparison with criteria

/compare #ClaimSetA #ClaimSetB --criteria scope,clarity

/test #idea

Validate idea in specific scenario

/test #Idea-24 --scenario "harsh environment"

/summarize [target]

Condense conversation or document

/summarize conversation --last 50

/critique [target]

Identify weaknesses and propose fixes

/critique #ClaimSetA --focus novelty,enablement

/scatter "concept"

Generate variants using innovation methods

/scatter "self-healing coating" --n=12 --methods=triz

/expand #idea

Deepen an idea with more detail

/expand Idea-17 --depth=2

/merge #idea1 #idea2

Synthesize multiple ideas into one

/merge Idea-12 Idea-17

/fork #idea

Create a derivative branch

/fork Idea-24 --reason "material change"

/cluster [items]

Group ideas by tags or similarity

/cluster Idea-* --by=embedding

/convert #document

Transform document to different format

/convert #Spec_v2 --to pdf

/evaluate #idea

Score against specific metrics

/evaluate Idea-24 --metrics novelty,feasibility,cost

/rank [items]

Order a list by criteria

/rank Idea-* --by=novelty

/validate #document

Check compliance against standard

/validate #ClaimSetA --against uspto-rules

/poll [options]

Start a vote with multiple options

/poll VariantA VariantB VariantC --method=borda

/vote [choice]

Cast a vote in an active poll

/vote Poll-12 VariantB

/draft #document

Generate content for document or section

/draft #Spec_v2.section(3) --style=uspto-plain

/revise #document

Improve existing document content

/revise #Abstract --focus clarity

/proofread #document

Check and repair text quality

/proofread #Spec_v2 --level=deep

/revert #document

Roll back to previous version

/revert #Spec_v2 --version v1.8

/assemble [parts]

Compile full package from parts

/assemble --parts claims,spec,figures

/save [item]

Store knowledge or preference

/save key:preferred-style value:uspto-plain

/assign [task] @agent

Delegate task to an agent

/assign draftClaims @claimsSpecialist --due 2025-10-10

/status [task]

Get progress report on task

/status task-123

/do [workflow]

Execute a standard workflow

/do priorArtSearch --scope core+adjacent

/find "expertise"

Discover agents by expertise or question

/find "who can explain CRISPR patent strategy?"

/add @agent

Bring agent into conversation

/add @examiner

/remove @agent

Remove agent from conversation

/remove @strategist

/mute @agent

Silence agent temporarily

/mute @virtuoso

/tune @agent

Configure specific agent behavior

/tune @examiner --style=concise --critique-depth=2

/prefs

Set overall conversation preferences

/prefs --length=brief --examples=always

/mode [type]

Switch collaboration mode

/mode roundtable

/style [preset]

Set writing style

/style patent-technical

/temperature [value]

Adjust AI creativity level

/temperature 0.7

Advanced Features

schedule

Automated Triggers

Set up conditional actions that execute automatically when conditions are met

/trigger onTimer 48h then /do review #Spec_v2
/trigger onDocumentComplete #Spec_v2 then /do priorArtSearch --scope core
/trigger onPollThreshold Poll123 0.66 then /merge top2
preview

Safe Testing

Preview command effects without making changes using dry-run mode

/merge #Idea12 #Idea17 --dryRun
→ Shows preview without executing
/prune score<0.5 --dryRun
→ Lists items to be removed

Pro Tip: Add --trace to any command to get a trace_id for audit logging

Master the Chat Interface

Start with natural conversation, then gradually adopt commands as you become comfortable. The interface grows with your expertise.