Autonomous Modes

Requires Login (Free+): Autonomous modes (/auto, /security, /monitor) require a free HelixMind account. Run helixmind login to unlock. it's free, no credit card needed. In Open Source mode, these commands are not available.

HelixMind can work autonomously in the background while you keep coding.

Autonomous Scan (/auto)

/auto

The agent works through issues in priority order:

Priority	Category
1	Critical bugs. crashes, null pointers, unhandled exceptions
2	Security vulnerabilities. injection, XSS, hardcoded secrets
3	Missing error handling. uncaught promises, unvalidated input
4	Broken tests. failing cases, outdated assertions
5	Performance bottlenecks. N+1 queries, unnecessary re-renders
6	Dead code. unused functions, unreachable branches
7	Code quality. linting issues, style inconsistencies

Goal-Driven Mode

/auto Refactor all API endpoints to use the new validation middleware

The agent breaks the goal into subtasks and works until completion.

Security Audit (/security)

/security

Run a one-time security audit of your codebase:

Category	What's Checked
Injection	SQL, command, template injection
Path Traversal	Directory traversal, symlink attacks
XSS	Cross-site scripting, unsanitized output
Secrets	Hardcoded API keys, passwords, tokens
Dependencies	Known vulnerabilities, outdated packages
Validation	Missing input validation, type coercion
Auth	Weak auth, missing session management
Race Conditions	TOCTOU bugs, concurrent access
Info Disclosure	Stack traces, debug info in production

Auto-Fix Behavior

Severity	Action
CRITICAL	Fixed immediately
HIGH	Fixed with notification
MEDIUM	Reported for manual review
LOW	Reported for manual review

Findings appear in the Brain's findings panel.

Continuous Security Monitor (/monitor)

/monitor [mode] [interval]

Start continuous security monitoring in the background:

Mode	Description
passive	Detects threats but takes no action (audit only)
defensive	Takes pre-approved defensive actions automatically
active	Requests approval for each defensive action via web dashboard

Examples:

/monitor                    # Passive mode, 5 min interval
/monitor active 60         # Active mode, 60 sec interval
/monitor defensive 180     # Defensive mode, 3 min interval

Features:

• Real-time threat detection (bruteforce, portscan, malware, config changes)
• Automated defenses (block IPs, kill processes, close ports)
• Web dashboard integration (/dashboard/monitor)
• Approval queue for manual review (Active mode)
• Baseline system scanning and drift detection

See the Security Monitor guide for full details.

Background Sessions

When you launch /auto or /security:

1. New session created with isolated state
2. You keep your interactive prompt
3. Background session shows as a tab
4. Notifications appear on completion
5. Sessions auto-close after 10 seconds

Each session has its own: conversation history, undo stack, agent controller.

Controls

Command	Action
/sessions	List all sessions
/stop	Stop background sessions
Ctrl+PageUp/Down	Switch tabs
ESC ESC	Emergency stop all

Tips

1. Run /security on every new project for a baseline assessment
2. Use /auto for cleanup while you focus on features
3. Review findings in the Brain before committing
4. Set specific goals with /auto for targeted work
5. Use checkpoints. /undo reverts unwanted changes