Security audit
Security checks for vulnerabilities and agentic risk
This weather skill uses external weather and geocoding services as expected, with an optional AirNow API key, and I found no hidden persistence, prompt hijacking, destructive behavior, or unrelated data access.
Install only if you are comfortable sending weather locations, including precise addresses or ZIP codes if you enter them, to external services. Configure AIRNOW_API_KEY only if you want AQI data and are comfortable with that key being used in AirNow API requests.
The declared purpose focuses on weather forecasting, but the skill also advertises AQI retrieval, astronomical calculations, aviation forecasts, and fire weather handling. This mismatch weakens informed consent and reviewability because operators may approve the skill for one purpose while it performs additional networked behaviors and processes extra categories of user queries.
The skill documentation indicates it uses environment variables and network access, but it does not declare any explicit tool scope such as permissions or allowed-tools. That creates an authorization gap where the runtime may grant broader capabilities than users or the platform expect, reducing transparency and making abuse harder to constrain.
The manifest describes a weather skill focused on US forecasts, accumulation data, warnings, and global fallback, but the module documentation explicitly adds aviation TAFs and fire weather as Phase 3 capabilities. Those specialized domains are not mentioned in the manifest description or intended-use text, so the implemented feature set exceeds the declared behavior.
User-supplied location queries are sent to third-party services (Nominatim and later wttr.in/NWS) without any visible user consent or disclosure mechanism in the skill. Because location data can be sensitive, especially precise addresses or ZIP codes, this creates a privacy risk through external transmission beyond the local agent environment.
Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.
def get_nws_gridpoint(lat, lon):
"""Get NWS gridpoint for lat/lon"""
try:
url = f"https://api.weather.gov/points/{lat},{lon}"
req = urllib.request.Request(url, headers={'User-Agent': 'ClawdWeather/1.0'})
with urllib.request.urlopen(req, timeout=10) as response:
Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.
def get_nws_gridpoint(lat, lon):
"""Get NWS gridpoint for lat/lon"""
try:
url = f"https://api.weather.gov/points/{lat},{lon}"
req = urllib.request.Request(url, headers={'User-Agent': 'ClawdWeather/1.0'})
with urllib.request.urlopen(req, timeout=10) as response:
Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.
def get_nws_gridpoint(lat, lon):
"""Get NWS gridpoint for lat/lon"""
try:
url = f"https://api.weather.gov/points/{lat},{lon}"
req = urllib.request.Request(url, headers={'User-Agent': 'ClawdWeather/1.0'})
with urllib.request.urlopen(req, timeout=10) as response:
Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.
def get_nws_gridpoint(lat, lon):
"""Get NWS gridpoint for lat/lon"""
try:
url = f"https://api.weather.gov/points/{lat},{lon}"
req = urllib.request.Request(url, headers={'User-Agent': 'ClawdWeather/1.0'})
with urllib.request.urlopen(req, timeout=10) as response:
Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.
def get_nws_gridpoint(lat, lon):
"""Get NWS gridpoint for lat/lon"""
try:
url = f"https://api.weather.gov/points/{lat},{lon}"
req = urllib.request.Request(url, headers={'User-Agent': 'ClawdWeather/1.0'})
with urllib.request.urlopen(req, timeout=10) as response:
Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.
def get_nws_gridpoint(lat, lon):
"""Get NWS gridpoint for lat/lon"""
try:
url = f"https://api.weather.gov/points/{lat},{lon}"
req = urllib.request.Request(url, headers={'User-Agent': 'ClawdWeather/1.0'})
with urllib.request.urlopen(req, timeout=10) as response:
Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.
def get_nws_gridpoint(lat, lon):
"""Get NWS gridpoint for lat/lon"""
try:
url = f"https://api.weather.gov/points/{lat},{lon}"
req = urllib.request.Request(url, headers={'User-Agent': 'ClawdWeather/1.0'})
with urllib.request.urlopen(req, timeout=10) as response:
The code accesses a sensitive environment variable, AIRNOW_API_KEY, and appends it to outbound HTTP requests. Although this is functionally expected for API usage, the file contains no visible warning in help text, comments for users, or other user-facing disclosure that a credential is consumed and sent to an external service.
The code accesses AIRNOW_API_KEY from environment variables to call a third-party air-quality API. While AQI support is documented in code, credential access is a broader capability than a straightforward public weather fetcher and is not justified by the manifest text, which does not mention secret handling or credentialed integrations.
No suspicious patterns detected.