Back to skill

Security audit

web-checker

Security checks for vulnerabilities and agentic risk

Overview

ClawScan rates this skill suspicious because it reports unrestricted redirects that could reach internal services and crawl and reporting weaknesses that may exceed or undermine its stated low-volume web-checking scope.

Consider the suspicious verdict and the reported redirect and crawl-limit weaknesses before installing. The scan inspected all 44 listed components, but it is marked partial because several references in SKILL.md could not be resolved; evidence is therefore incomplete. The findings do not establish malicious intent, credential theft, or data exfiltration.

Vulnerability Patterns
  • Insecure Skill Coding PracticesFinds exploitable flaws such as hardcoded secrets or command injection
  • Skill Instruction HijackingAlters the agent's session goals or safety constraints when the skill loads
  • Agent Memory PoisoningWrites attacker-controlled rules into memory that affect later sessions
  • Remote Payload Retrieval and ExecutionFetches external code whose behavior can change after review
  • Embedded Malicious CodeShips malicious scripts inside the skill and executes them locally
Findings (1)

T09 · Insecure Skill Coding Practices

Warning
Location
recon-cli/recon_cli/http.py
Finding

Unrestricted redirects permit requests to internal or out-of-scope services

Content
View full analysis

Impact

An attacker can induce HTTP requests to internal or otherwise unauthorized services. When the HTML-shell check receives a readable internal HTML response, extracted form tags and comments can also appear in the local audit report. The demonstrated capability does not establish credential theft, external exfiltration, or code execution.

Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Behavioral ASTexec() Call, eval() Call, Dynamic Import
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Findings (27)

Credential Access

High
Category
Privilege Escalation
Confidence
60% confidence
Finding

Code accesses credential files (SSH keys, AWS credentials, etc.). This could indicate credential theft attempts.

Content

Scanner excerpt · recon-cli/build/lib/recon_cli/checks/sensitive_paths.py (reported line 41)May include surrounding context.

python
# (path, description, severity-if-truly-live)
PATHS: List[Tuple[str, str, Severity]] = [
    (".env",                    "Env vars (real secrets)",        Severity.CRIT),
    (".env.local",              "Local env",                      Severity.CRIT),
    (".env.production",         "Production env",                 Severity.CRIT),
    (".git/config",             "Git repo (source leak)",         Severity.CRIT),

Credential Access

High
Category
Privilege Escalation
Confidence
60% confidence
Finding

Code accesses credential files (SSH keys, AWS credentials, etc.). This could indicate credential theft attempts.

Content

Scanner excerpt · recon-cli/build/lib/recon_cli/checks/sensitive_paths.py (reported line 66)May include surrounding context.

python
# The SECRET, API_KEY, DATABASE_URL, PRIVATE_KEY, AWS_, OPENAI_ tokens are
# common prefixes in public .env exposures — included here as regex fragments.
MARKERS = {
    ".env":            [r"^\s*(SECRET|API_KEY|DATABASE_URL|PRIVATE_KEY|AWS_|OPENAI_)\s*=", r"BEGIN (RSA|EC|DSA|OPENSSH) PRIVATE KEY"],
    ".git/config":     [r"^\[core\]$", r"^\s*(remote|branch|worktree)\s*="],
    ".git/HEAD":       [r"^ref:\s*refs/heads/", r"^[0-9a-f]{40}\s*$"],
    "wp-login.php":    [r"wp-login\.php", r"wp\.com|WordPress"],

Credential Access

High
Category
Privilege Escalation
Confidence
60% confidence
Finding

Code accesses credential files (SSH keys, AWS credentials, etc.). This could indicate credential theft attempts.

Content

Scanner excerpt · recon-cli/recon_cli/checks/sensitive_paths.py (reported line 41)May include surrounding context.

python
# (path, description, severity-if-truly-live)
PATHS: List[Tuple[str, str, Severity]] = [
    (".env",                    "Env vars (real secrets)",        Severity.CRIT),
    (".env.local",              "Local env",                      Severity.CRIT),
    (".env.production",         "Production env",                 Severity.CRIT),
    (".git/config",             "Git repo (source leak)",         Severity.CRIT),

Credential Access

High
Category
Privilege Escalation
Confidence
60% confidence
Finding

Code accesses credential files (SSH keys, AWS credentials, etc.). This could indicate credential theft attempts.

Content

Scanner excerpt · recon-cli/recon_cli/checks/sensitive_paths.py (reported line 66)May include surrounding context.

python
# The SECRET, API_KEY, DATABASE_URL, PRIVATE_KEY, AWS_, OPENAI_ tokens are
# common prefixes in public .env exposures — included here as regex fragments.
MARKERS = {
    ".env":            [r"^\s*(SECRET|API_KEY|DATABASE_URL|PRIVATE_KEY|AWS_|OPENAI_)\s*=", r"BEGIN (RSA|EC|DSA|OPENSSH) PRIVATE KEY"],
    ".git/config":     [r"^\[core\]$", r"^\s*(remote|branch|worktree)\s*="],
    ".git/HEAD":       [r"^ref:\s*refs/heads/", r"^[0-9a-f]{40}\s*$"],
    "wp-login.php":    [r"wp-login\.php", r"wp\.com|WordPress"],

Credential Access

High
Category
Privilege Escalation
Confidence
60% confidence
Finding

Code accesses credential files (SSH keys, AWS credentials, etc.). This could indicate credential theft attempts.

Content

Scanner excerpt · recon-cli/build/lib/recon_cli/checks/sensitive_paths.py (reported line 42)May include surrounding context.

python
# (path, description, severity-if-truly-live)
PATHS: List[Tuple[str, str, Severity]] = [
    (".env",                    "Env vars (real secrets)",        Severity.CRIT),
    (".env.local",              "Local env",                      Severity.CRIT),
    (".env.production",         "Production env",                 Severity.CRIT),
    (".git/config",             "Git repo (source leak)",         Severity.CRIT),
    (".git/HEAD",               "Git repo (source leak)",         Severity.CRIT),

Credential Access

High
Category
Privilege Escalation
Confidence
60% confidence
Finding

Code accesses credential files (SSH keys, AWS credentials, etc.). This could indicate credential theft attempts.

Content

Scanner excerpt · recon-cli/recon_cli/checks/sensitive_paths.py (reported line 42)May include surrounding context.

python
# (path, description, severity-if-truly-live)
PATHS: List[Tuple[str, str, Severity]] = [
    (".env",                    "Env vars (real secrets)",        Severity.CRIT),
    (".env.local",              "Local env",                      Severity.CRIT),
    (".env.production",         "Production env",                 Severity.CRIT),
    (".git/config",             "Git repo (source leak)",         Severity.CRIT),
    (".git/HEAD",               "Git repo (source leak)",         Severity.CRIT),

Credential Access

High
Category
Privilege Escalation
Confidence
60% confidence
Finding

Code accesses credential files (SSH keys, AWS credentials, etc.). This could indicate credential theft attempts.

Content

Scanner excerpt · recon-cli/build/lib/recon_cli/checks/sensitive_paths.py (reported line 43)May include surrounding context.

python
PATHS: List[Tuple[str, str, Severity]] = [
    (".env",                    "Env vars (real secrets)",        Severity.CRIT),
    (".env.local",              "Local env",                      Severity.CRIT),
    (".env.production",         "Production env",                 Severity.CRIT),
    (".git/config",             "Git repo (source leak)",         Severity.CRIT),
    (".git/HEAD",               "Git repo (source leak)",         Severity.CRIT),
    ("wp-login.php",            "WordPress login (if it exists)", Severity.HIGH),

Credential Access

High
Category
Privilege Escalation
Confidence
60% confidence
Finding

Code accesses credential files (SSH keys, AWS credentials, etc.). This could indicate credential theft attempts.

Content

Scanner excerpt · recon-cli/recon_cli/checks/sensitive_paths.py (reported line 43)May include surrounding context.

python
PATHS: List[Tuple[str, str, Severity]] = [
    (".env",                    "Env vars (real secrets)",        Severity.CRIT),
    (".env.local",              "Local env",                      Severity.CRIT),
    (".env.production",         "Production env",                 Severity.CRIT),
    (".git/config",             "Git repo (source leak)",         Severity.CRIT),
    (".git/HEAD",               "Git repo (source leak)",         Severity.CRIT),
    ("wp-login.php",            "WordPress login (if it exists)", Severity.HIGH),

Credential Access

High
Category
Privilege Escalation
Confidence
60% confidence
Finding

Code accesses credential files (SSH keys, AWS credentials, etc.). This could indicate credential theft attempts.

Content

Scanner excerpt · recon-cli/build/lib/recon_cli/checks/sensitive_paths.py (reported line 64)May include surrounding context.

python
# NOTE: These are *pattern strings* used to detect exposed .env-style files
# on remote web servers.  They are NOT real credential values or references.
# The SECRET, API_KEY, DATABASE_URL, PRIVATE_KEY, AWS_, OPENAI_ tokens are
# common prefixes in public .env exposures — included here as regex fragments.
MARKERS = {
    ".env":            [r"^\s*(SECRET|API_KEY|DATABASE_URL|PRIVATE_KEY|AWS_|OPENAI_)\s*=", r"BEGIN (RSA|EC|DSA|OPENSSH) PRIVATE KEY"],
    ".git/config":     [r"^\[core\]$", r"^\s*(remote|branch|worktree)\s*="],

Credential Access

High
Category
Privilege Escalation
Confidence
60% confidence
Finding

Code accesses credential files (SSH keys, AWS credentials, etc.). This could indicate credential theft attempts.

Content

Scanner excerpt · recon-cli/recon_cli/checks/sensitive_paths.py (reported line 64)May include surrounding context.

python
# NOTE: These are *pattern strings* used to detect exposed .env-style files
# on remote web servers.  They are NOT real credential values or references.
# The SECRET, API_KEY, DATABASE_URL, PRIVATE_KEY, AWS_, OPENAI_ tokens are
# common prefixes in public .env exposures — included here as regex fragments.
MARKERS = {
    ".env":            [r"^\s*(SECRET|API_KEY|DATABASE_URL|PRIVATE_KEY|AWS_|OPENAI_)\s*=", r"BEGIN (RSA|EC|DSA|OPENSSH) PRIVATE KEY"],
    ".git/config":     [r"^\[core\]$", r"^\s*(remote|branch|worktree)\s*="],

subprocess module call

Medium
Category
Dangerous Code Execution
Confidence
70% confidence
Finding

subprocess module calls execute external commands. Without careful input validation, this enables command injection.

Content

Scanner excerpt · recon-cli/build/lib/recon_cli/checks/dns.py (reported line 24)May include surrounding context.

python
if not host or not _SAFE_DIG_HOST.fullmatch(host):
        return None
    try:
        p = subprocess.run(
            ["dig", "+short", host, rtype],
            capture_output=True,
            text=True,

subprocess module call

Medium
Category
Dangerous Code Execution
Confidence
70% confidence
Finding

subprocess module calls execute external commands. Without careful input validation, this enables command injection.

Content

Scanner excerpt · recon-cli/recon_cli/checks/dns.py (reported line 24)May include surrounding context.

python
if not host or not _SAFE_DIG_HOST.fullmatch(host):
        return None
    try:
        p = subprocess.run(
            ["dig", "+short", host, rtype],
            capture_output=True,
            text=True,

External Transmission

Medium
Category
Data Exfiltration
Confidence
70% confidence
Finding

Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Content

Scanner excerpt · recon-cli/build/lib/recon_cli/checks/headers.py (reported line 20)May include surrounding context.

python
def _parse_headers(raw: str) -> Dict[str, str]:
    """curl -D - output may include a status line; keep last header block."""
    # take the last HTTP/xxx status section if -L followed redirects
    blocks = re.split(r"\n\n(?=HTTP/)", _normalise(raw))
    last = blocks[-1] if blocks else raw

External Transmission

Medium
Category
Data Exfiltration
Confidence
70% confidence
Finding

Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Content

Scanner excerpt · recon-cli/recon_cli/checks/headers.py (reported line 20)May include surrounding context.

python
def _parse_headers(raw: str) -> Dict[str, str]:
    """curl -D - output may include a status line; keep last header block."""
    # take the last HTTP/xxx status section if -L followed redirects
    blocks = re.split(r"\n\n(?=HTTP/)", _normalise(raw))
    last = blocks[-1] if blocks else raw

The JS crawl can make unbounded download attempts despite the skill's low-volume scope.

Medium
Category
Excessive Agency
Confidence
99% confidence
Finding

The loop limits successfully loaded bundles rather than total download attempts, so failed downloads do not consume the configured budget. An adversarial shell or bundle can supply many distinct failing references, causing excessive requests and prolonged execution despite the advertised low-volume scope.

Content

No source excerpt is available for this finding.

subprocess module call

Medium
Category
Dangerous Code Execution
Confidence
70% confidence
Finding

subprocess module calls execute external commands. Without careful input validation, this enables command injection.

Content

Scanner excerpt · recon-cli/build/lib/recon_cli/checks/tls.py (reported line 25)May include surrounding context.

python
if not isinstance(port, int) or not 0 < port < 65536:
        return None
    try:
        p = subprocess.run(
            ["openssl", "s_client", "-connect", f"{host}:{port}", "-servername", host],
            input=b"\n",
            capture_output=True,

subprocess module call

Medium
Category
Dangerous Code Execution
Confidence
70% confidence
Finding

subprocess module calls execute external commands. Without careful input validation, this enables command injection.

Content

Scanner excerpt · recon-cli/recon_cli/checks/tls.py (reported line 25)May include surrounding context.

python
if not isinstance(port, int) or not 0 < port < 65536:
        return None
    try:
        p = subprocess.run(
            ["openssl", "s_client", "-connect", f"{host}:{port}", "-servername", host],
            input=b"\n",
            capture_output=True,

subprocess module call

Medium
Category
Dangerous Code Execution
Confidence
70% confidence
Finding

subprocess module calls execute external commands. Without careful input validation, this enables command injection.

Content

Scanner excerpt · recon-cli/build/lib/recon_cli/checks/tls.py (reported line 39)May include surrounding context.

python
def _x509(s_client_out: str, cfg: Config) -> Optional[str]:
    """Extract subject/issuer/dates/SAN from openssl s_client output."""
    try:
        p = subprocess.run(
            ["openssl", "x509", "-noout", "-subject", "-issuer", "-dates",
             "-ext", "subjectAltName"],
            input=s_client_out.encode("utf-8"),

subprocess module call

Medium
Category
Dangerous Code Execution
Confidence
70% confidence
Finding

subprocess module calls execute external commands. Without careful input validation, this enables command injection.

Content

Scanner excerpt · recon-cli/recon_cli/checks/tls.py (reported line 39)May include surrounding context.

python
def _x509(s_client_out: str, cfg: Config) -> Optional[str]:
    """Extract subject/issuer/dates/SAN from openssl s_client output."""
    try:
        p = subprocess.run(
            ["openssl", "x509", "-noout", "-subject", "-issuer", "-dates",
             "-ext", "subjectAltName"],
            input=s_client_out.encode("utf-8"),

subprocess module call

Medium
Category
Dangerous Code Execution
Confidence
70% confidence
Finding

subprocess module calls execute external commands. Without careful input validation, this enables command injection.

Content

Scanner excerpt · recon-cli/build/lib/recon_cli/checks/tls.py (reported line 55)May include surrounding context.

python
def _signature(s_client_out: str, cfg: Config) -> Optional[str]:
    """Read the signature algorithm (not included in the short x509 dump)."""
    try:
        p = subprocess.run(
            ["openssl", "x509", "-noout", "-text"],
            input=s_client_out.encode("utf-8"),
            capture_output=True,

subprocess module call

Medium
Category
Dangerous Code Execution
Confidence
70% confidence
Finding

subprocess module calls execute external commands. Without careful input validation, this enables command injection.

Content

Scanner excerpt · recon-cli/recon_cli/checks/tls.py (reported line 55)May include surrounding context.

python
def _signature(s_client_out: str, cfg: Config) -> Optional[str]:
    """Read the signature algorithm (not included in the short x509 dump)."""
    try:
        p = subprocess.run(
            ["openssl", "x509", "-noout", "-text"],
            input=s_client_out.encode("utf-8"),
            capture_output=True,

External Transmission

Medium
Category
Data Exfiltration
Confidence
70% confidence
Finding

Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Content

Scanner excerpt · recon-cli/build/lib/recon_cli/http.py (reported line 3)May include surrounding context.

python
"""HTTP helpers built on `curl` (with a stdlib fallback).

We prefer curl because it already handles:
  - redirects (-L), HTTP/2, SNI, cert verification
  - clean header dumps (-I)
  - fast, low-overhead requests

External Transmission

Medium
Category
Data Exfiltration
Confidence
70% confidence
Finding

Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Content

Scanner excerpt · recon-cli/recon_cli/http.py (reported line 3)May include surrounding context.

python
"""HTTP helpers built on `curl` (with a stdlib fallback).

We prefer curl because it already handles:
  - redirects (-L), HTTP/2, SNI, cert verification
  - clean header dumps (-I)
  - fast, low-overhead requests

subprocess module call

Medium
Category
Dangerous Code Execution
Confidence
70% confidence
Finding

subprocess module calls execute external commands. Without careful input validation, this enables command injection.

Content

Scanner excerpt · recon-cli/build/lib/recon_cli/http.py (reported line 45)May include surrounding context.

python
*args,
    ]
    try:
        p = subprocess.run(
            cmd, capture_output=True, text=True, timeout=cfg.timeout + 5
        )
    except Exception as e:  # noqa: BLE001

SPA catch-all handling claims shell identity without comparing response bodies.

Medium
Category
Excessive Agency
Confidence
99% confidence
Finding

HTML responses without recognized path-specific markers are classified as SPA fallbacks without comparing their bodies to the root response, yet the report claims they are identical. Distinct administrative pages or HTML-formatted sensitive disclosures can therefore be suppressed, creating false assurance in a tool whose purpose is detecting these exposures.

Content

No source excerpt is available for this finding.