NeuralCodeDOCS
Configuration

Agents & pipelines

An agent is a named instruction with its own model, thinking budget, and permission mode. Select a few from the picker and your task runs through them in order — a pipeline, where each stage hands off to the next.

NeuralCode
Agents— select a pipeline, runs top-to-bottom
1/planMap the work before touching codeplan-modeOpus
2/reviewFind bugs, risks & smellsread-onlyOpus
3/testWrite and run testsauto-editSonnet
/goalClarify the objectiveaskOpus
/optimizeSpeed & memoryauto-editOpus
/refactorClean up, same behaviorauto-editSonnet
/docsWrite the docsauto-editHaiku
/securityAudit for vulnerabilitiesread-onlyOpus
/typefixResolve type errorsauto-editSonnet
/prOpen a pull requestfull-accessSonnet
Refactor the auth module and ship a PR…
Agents1plan2review3test
Pipeline3 stages
1/plan
plan-mode · Opus
2/review
read-only · Opus
3/testqueued
auto-edit · Sonnet
each stage carries its own model · thinking · permission
The 🧠 Agents picker on the left; the selected agents become an ordered pipeline that runs top-to-bottom.

The ten built-in agents#

NeuralCode ships with ten agents covering the work you do most. Each one comes pre-tuned with a sensible model, thinking budget, and permission mode — for example, /plan runs in plan-mode so it maps the work without touching files, and /review is read-only.

AgentWhat it doesPermission
/planMaps the work before any code changes.plan-mode
/goalClarifies the objective and success criteria.ask
/reviewFinds bugs, risks, and code smells.read-only
/testWrites and runs tests for the change.auto-edit
/optimizeImproves speed and memory.auto-edit
/refactorCleans up without changing behavior.auto-edit
/docsWrites or updates documentation.auto-edit
/securityAudits the change for vulnerabilities.read-only
/typefixResolves type errors.auto-edit
/prOpens a pull request for the work.full-access

The composer picker#

Open the 🧠 Agents button in the composer to browse and select agents. Selected agents show as numbered chips — the numbers are the run order. You can also just type the name (/plan, /review, …) and it's selected for you, the same way commands & skills work.

Pipeline execution#

When you send the task, it runs through the selected agents in order. Each stage executes with its own settings — so a pipeline might plan in Opus with extended thinking under plan-mode, review read-only, then write tests with auto-edit on Sonnet. The output of one stage is the starting context for the next.

  1. 1
    Pick the agents
    Click 🧠 Agents (or type their names) to build the chip row — e.g. /plan/review /test.
  2. 2
    Describe the task
    Write your prompt as usual. It's fed into the first agent.
  3. 3
    Watch it flow
    Each stage runs in order with a live status; finished stages show a check, the current one spins, the rest are queued.
Order matters
Put planning and clarification first, edits in the middle, and review or /pr last. A read-only /review after your edits catches problems before they reach you.

Settings → Agents#

Manage agents in Settings → Agents. Enable or disable any agent, edit its prompt, model, thinking budget, and permission mode — or add your own and delete the custom ones you no longer need.

Back to app
General
Models
Agents
Workflows
MCP servers
Commands & Skills
Appearance
Git
Account
About
v0.1.0-beta.1

Agents

Enable, edit, or add your own. Each agent is a named instruction with its own model, thinking budget, and permission mode.

/plan
/review
/test
/goal
/optimize
/refactor
Add agent
/review
Model
Opus 4.8
Thinking
On
Permission
read-only
Prompt
Review the diff for correctness bugs and security issues. Do not edit any files — only report findings, grouped by severity, and end with a one-line verdict.
Settings → Agents: toggle agents on or off and edit the prompt, model, thinking, and permission for each.
Note
Custom agents behave exactly like built-ins — they appear in the picker, can be typed by name, and slot into a pipeline anywhere you put them.
Reuse a pipeline
Run the same sequence often? Save it as a workflow and apply the whole pipeline from the composer in one click.