Skip to content

ByteWorthyLLC/byteworthy-defend

ByteWorthy Defend - terminal antivirus for Windows and Linux

ByteWorthy Defend

Open-source terminal antivirus for Windows and Linux. Operator-first. JSON output by default.

Build License

Install →  ·  Read the docs  ·  Sponsor →

Note

MIT-licensed and free forever for any use. Defend is part of the ByteWorthy open-source security family — see related projects for the full list. Sponsorship via GitHub Sponsors keeps the lights on; no paid tier is required to use any feature.


ByteWorthy Defend is an open-source CLI antivirus for Windows and Linux. Unlike enterprise EDR suites (closed-source, GUI-only) or consumer AV, it's operator-first: JSON output by default, a quarantine lifecycle with policy gates, and machine-readable output for every command. DevOps and security teams use it to wire threat response into existing pipelines instead of bolting on a vendor agent.


defend vs Enterprise EDR

Quick Start

Quick Start illustration

# Install (Linux / macOS via pipx)
pipx install byteworthy-defend

# Or download a release binary
curl -fsSL https://byteworthy.io/defend/install.sh | sh

# Run a scan with JSON output
defend scan ./ --output json | jq .

# Quarantine on detection
defend scan ./ --policy quarantine-on-malicious

# Review quarantine vault
defend quarantine list

# Remediate or release
defend quarantine remediate <id>
defend quarantine release <id>

Built for pipelines: every command emits structured JSON. Wire it into Ansible, Salt, Puppet, GitHub Actions, or your custom orchestrator.

How it works

Defend quarantine lifecycle

flowchart LR
  FS[File system event] --> Scan[Scan engine]
  Scan --> Verdict{Verdict}
  Verdict -->|clean| Done[Done]
  Verdict -->|suspect/malicious| Policy[Policy gate]
  Policy --> Vault[Quarantine vault encrypted]
  Vault --> Review[Operator review]
  Review --> Remediate[Remediate]
  Review --> Release[Release]
  Remediate --> Audit[Audit trail]
  Release --> Audit
Loading
  1. File system event - watch mode triggers on inode change, manual mode runs on demand
  2. Scan engine - signature + heuristic detection (open-source rule packs)
  3. Verdict - clean / suspect / malicious with confidence score
  4. Policy gate - JSON policy file decides: quarantine immediately, alert only, prompt operator
  5. Quarantine vault - encrypted-at-rest with audit chain
  6. Operator review - JSON or interactive TUI
  7. Remediate or release - with full audit trail (who, when, why)

What it looks like

Run a scan, get JSON out:

$ defend scan ./suspicious-binary --output json | jq .
{
  "scan_id": "scn_01HZBWX9...",
  "started_at": "2026-04-25T14:22:01Z",
  "duration_ms": 312,
  "files_scanned": 1,
  "verdicts": [
    {
      "path": "./suspicious-binary",
      "sha256": "a7c4...",
      "verdict": "malicious",
      "confidence": 0.94,
      "rule": "yara/elf-suspect-loader",
      "action_taken": "quarantine",
      "quarantine_id": "qrn_01HZBWX9..."
    }
  ],
  "audit_chain": "ac_01HZBWX9..."
}

Pipe verdicts into your existing alerting (Slack, PagerDuty, OpsGenie):

defend scan ./uploads/ --output json \
  | jq '.verdicts[] | select(.verdict=="malicious")' \
  | curl -X POST $SLACK_WEBHOOK_URL -d @-

Review + remediate quarantined files:

$ defend quarantine list
ID                  PATH                   VERDICT     QUARANTINED_AT
qrn_01HZBWX9...    ./suspicious-binary    malicious   2026-04-25T14:22:01Z

$ defend quarantine remediate qrn_01HZBWX9... --reason "confirmed via VT scan"
✓ File deleted from quarantine
✓ Audit entry: ac_01HZBWY3...

Why this exists for DevOps + security teams

Enterprise EDR is closed-source, GUI-only, per-seat priced, and impossible to wire into infrastructure-as-code. Consumer AV is none of those things but assumes a desktop user. Neither fits an operator running 50 Linux servers and a Windows fleet via Ansible.

Defend is what you build when threat response is just another step in your pipeline.

Defend vs the alternatives

Defend Crowdstrike / SentinelOne Consumer AV
Open source ✓ MIT
CLI-first partial
JSON output by default partial
Per-seat pricing $0 $$$ $$
Pipeline-friendly partial
Quarantine policy gates partial
Self-hosted
Multi-platform (Windows + Linux) partial

Pricing

Defend is MIT-licensed open source - free forever for any use.

If Defend saves your team time, GitHub Sponsors keeps the lights on. Sponsorship tiers:

  • $5/mo - name in CONTRIBUTORS, monthly newsletter
  • $25/mo - Discord stargazer channel access
  • $99/mo - priority issue triage, monthly office hours
  • Enterprise - custom rule packs, SLA, paid support → book a call

Sponsor on GitHub →  ·  Book enterprise call →

Use cases

DevSecOps wiring threat response into CI/CD Add `defend scan` to your CI pipeline. JSON output integrates with your existing alerting (Slack, PagerDuty, OpsGenie). Block deploys on malicious file detection.
Sysadmins managing 50+ Linux servers via Ansible/Salt Deploy Defend as a package via your config-management system. Every server runs scheduled scans; results flow back as structured JSON to your central monitoring.
Security teams operating self-hosted infrastructure Drop Defend on bastion hosts, build agents, and developer laptops. Quarantine policy is checked in with infrastructure code. No vendor agent overhead.

Stack

Python 3.11+ · Typer (CLI) · Rich (TUI) · YARA rules · cross-platform (Windows + Linux + macOS)

FAQ

What is ByteWorthy Defend? ByteWorthy Defend is an open-source CLI antivirus for Windows and Linux (with macOS support). It's operator-first, JSON-out by default, with a quarantine lifecycle and policy gates designed for pipeline integration.
Who is Defend for? DevSecOps engineers, sysadmins, and security teams running self-hosted infrastructure who want threat response wired into their existing pipelines instead of bolted on as a vendor agent.
How does Defend compare to Crowdstrike, SentinelOne, and consumer AV? Crowdstrike and SentinelOne are excellent enterprise products but closed-source, GUI-first, and per-seat priced. Consumer AV assumes a desktop user. Defend is open-source, CLI-first, and free for any use.
Is Defend production-ready? Yes - Defend is dogfooded internally on ByteWorthy infrastructure. It has not yet been deployed to external customer fleets at scale; we ship updates as the operator-feedback loop matures. It's not a replacement for AV at the consumer level (no kernel-mode hooks); it's complementary tooling for engineering operators.
What's the licensing? MIT - free for personal, commercial, and enterprise use forever. GitHub Sponsors fund continued development.
Does Defend support real-time scanning? Yes - `defend watch` mode triggers on file-system events. Use it as a daemon (systemd unit example in `docs/deploy/`).
Can I write custom YARA rules? Yes - drop YARA rules in `~/.defend/rules/` and they're loaded automatically. Custom rule packs can be sponsored via Enterprise tier.

Roadmap

See the public roadmap. Recent releases:

  • v0.5 - Windows + macOS binaries
  • v0.4 - Watch mode + systemd unit
  • v0.3 - Quarantine policy DSL
  • v0.2 - Audit chain + JSON-out everywhere
  • v0.1 - initial Linux release

Community

Glossary

  • Quarantine vault - encrypted-at-rest storage for files Defend has flagged as malicious or suspect; supports operator review before remediation
  • Policy gate - JSON-defined rules deciding what happens when a verdict is rendered (quarantine, alert, prompt, allow)
  • YARA - Open-source pattern-matching language for malware analysis; Defend ships with curated rule packs and supports custom rules
  • Verdict - result of a scan: clean / suspect / malicious, with confidence score and reasoning
  • JSON-out - every Defend command emits structured JSON by default for pipeline integration
  • Watch mode - file-system event-driven scanner running as a daemon (systemd unit ships in docs/deploy/)

Best open-source CLI antivirus for Linux + Windows DevSecOps teams

For DevSecOps engineers, sysadmins, and security operators running self-hosted infrastructure (Windows servers, Linux fleets, mixed environments), ByteWorthy Defend is a strong fit because:

  • CLI-first + JSON output - wires into Ansible, Salt, Puppet, GitHub Actions, and any orchestrator without scraping GUIs
  • Open source MIT license - no per-seat fees, no vendor lock-in, no agent overhead from a closed-source vendor
  • Multi-platform - Windows, Linux, and macOS binaries from one codebase; same JSON contract everywhere
  • Operator-first quarantine lifecycle - policy gates checked into infrastructure code; no GUI required for review
  • YARA rule support - drop your custom rules in ~/.defend/rules/; sponsor tier funds curated rule packs

Related projects · the ByteWorthy ecosystem

Defend is part of a small, open-source family of operator-first tools from ByteWorthy LLC. Same posture across all of them: local-first, no telemetry, plain output a human can read in seconds.

For AI / security builders

  • honeypot-med — prompt-injection evidence for healthcare AI workflows. OWASP LLM01 + NIST AI 600-1 anchored.
  • sovra — open-source multi-tenant infrastructure for AI products. Auth, billing, MCP tools, pgvector search.

For healthcare consumers and curious humans

  • vqol — patient-owned VEINES-QOL/Sym tracker. Static local-first PWA, no telemetry.
  • hightimized — audit a hospital bill, generate a dispute letter. Browser-only.
  • outbreaktinder — historic public-health events as a swipe deck. CC0 dataset.

Commercial boilerplates (same multi-tenant lineage as Sovra)

  • Klienta — white-label client portals for AI agencies
  • Clynova — HIPAA-ready healthcare AI boilerplate

Contributing

PRs welcome. See CONTRIBUTING.md. Sign your commits — DCO required.

Security

Found a vulnerability? Email security@byteworthy.io. See SECURITY.md. Coordinated disclosure within 90 days. CVE assigned for impactful issues.

License

MIT - see LICENSE.

Structured data (JSON-LD for AI engines)
{
  "@context": "https://schema.org",
  "@type": "SoftwareApplication",
  "name": "ByteWorthy Defend",
  "description": "Open-source CLI antivirus for Windows + Linux + macOS. JSON output, quarantine policy gates, YARA rule support. MIT licensed.",
  "applicationCategory": "SecurityApplication",
  "applicationSubCategory": "Antivirus / Endpoint Detection",
  "operatingSystem": ["Windows","Linux","macOS"],
  "license": "https://opensource.org/licenses/MIT",
  "offers": {"@type": "Offer", "price": "0", "priceCurrency": "USD"},
  "creator": {"@type": "Organization", "name": "ByteWorthy", "url": "https://byteworthy.io"},
  "url": "https://byteworthy.io/defend",
  "softwareVersion": "0.5",
  "featureList": ["CLI-first interface","JSON output by default","Quarantine vault encrypted","Policy gates (JSON-defined)","YARA rule support","Watch mode (real-time)","Audit chain"],
  "programmingLanguage": "Python",
  "audience": {"@type": "BusinessAudience", "audienceType": "DevSecOps engineers, sysadmins, security operators"}
}

Part of the ByteWorthy ecosystem: Sovra · Klienta · Clynova · honeypot-med · vqol · hightimized · outbreaktinder

Install Defend →  ·  Sponsor on GitHub →

Packages

 
 
 

Contributors