Back to skill

Security audit

Open-Meteo Weather + Weather Strip

Security checks for vulnerabilities and agentic risk

Overview

The skill is a coherent weather tool, but its generated HTML widget can preserve unsafe schedule names as executable browser content.

Review before installing. The basic weather lookup behavior is proportionate, but do not use untrusted or user-supplied schedule names with the weather strip generator until output escaping is fixed. Treat generated weather-strip HTML as active content, and prefer controlled output locations plus clear disclosure that coordinates are sent to Open-Meteo.

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

Error
Location
scripts/weather_strip.py:515
Finding

Stored HTML, SVG, and JavaScript Injection Through Unescaped Schedule Names

Content
View full analysis
0: elements.append( f'' ) elements.append( f'{hourly_cities[i]}' ) ``` ```python # scripts/weather_strip.py:593-601 scrubber_data.append({ "time": t.strftime("%-I:%M %p"), "day": t.strftime("%a %-m/%-d"), "temp": f"{temps[i]:.0f}{unit_sym}", "feels": f"{feels[i]:.0f}°", "dew": f"{dews[i]:.0f}°", "cloud": f"{clouds[i]}%", "rain": f"{rain_probs[i]}%", "wind": f"{winds[i]:.0f} {wind_unit}", "uv": f"{uv_vals[i]:.0f}", "humidity": f"{humids[i]}%", "city": hourly_cities[i] if i < len(hourly_cities) else "", }) ``` ```python # scripts/weather_strip.py:638-642 city = daily_cities[i] if i < len(daily_cities) else "" dsvg.append(f'{d.strftime("%a")}') if city and (i == 0 or daily_cities[i] != daily_cities[i-1]): dsvg.append(f'{city}') ``` ```python # scripts/weather_strip.py:858-887 const da ...[truncated 4701 chars]
Remediation
View remediation
`. Prefer a non-executable JSON container: ```html ``` Escape at least `<`, `>`, `&`, U+2028, and U+2029 before embedding, then parse the element's `textContent` with `JSON.parse()`. 3. Replace `innerHTML` with DOM construction and `textContent`. For example: ```javascript const cityNode = document.createElement("div"); cityNode.className = "ws-tt-city"; cityNode.textContent = d.city; tooltip.replaceChildren(cityNode); ``` Construct the remaining tooltip rows through `createElement()` and assign all data values through `textContent`. 4. Validate schedule names before processing. Apply a reasonable maximum length and, where compatible with product requirements, a conservative allowlist of letters, numbers, spaces, apostrophes, hyphens, and periods. 5. Apply a restrictive Content Security Policy to pages embedding the widget. Avoid `unsafe-inline` scripts and use nonces or external scripts. This is defense in depth and does not replace proper output encoding. 6. Add regression tests using city names containing: - `<`, `>`, `&`, single quotes, and double quotes. - Closing SVG or HTML tags. - `` sequences. - Event-handler attributes and SVG active-content payloads. Tests should verify that generated output displays these values strictly as text and never creates additional DOM elements or executable script. ]]>
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (8)

Ae1

High
Category
analysis-evasion
Confidence
100% confidence
Finding

Referenced artifact was not completely inspected

Content

Scanner excerpt · SKILL.md (reported line 45)May include surrounding context.

md
python3 scripts/weather_strip.py --lat 37.75 --lon -122.43 --days 7 \

Ae1

High
Category
analysis-evasion
Confidence
100% confidence
Finding

Referenced artifact was not completely inspected

Content

Scanner excerpt · SKILL.md (reported line 49)May include surrounding context.

md
python3 scripts/weather_strip.py --lat 37.75 --lon -122.43 --days 7 \

Ae1

High
Category
analysis-evasion
Confidence
100% confidence
Finding

Referenced artifact was not completely inspected

Content

Scanner excerpt · SKILL.md (reported line 53)May include surrounding context.

md
python3 scripts/weather_strip.py --lat 37.75 --lon -122.43 --days 7 \

Undeclared Tool Scope

Medium
Category
MCP Least Privilege
Confidence
92% confidence
Finding

The skill documents capabilities that imply shell execution, network access, and file writes, but it does not declare any explicit tool scope or permissions boundaries. That creates a least-privilege and governance gap: a host agent may grant broader access than users expect, and the examples include writing output to absolute local filesystem paths, which increases risk if the skill is invoked in a more privileged environment.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
92% confidence
Finding

The manifest says to use the skill whenever a user asks about weather, temperature, rain probability, UV index, wind, or forecasts for any location. This trigger scope is broad and lacks constraints or exclusion examples, so it could cause unintended invocation during general conversation that mentions temperature or wind without actually requesting a weather lookup.

Content

No source excerpt is available for this finding.

External Transmission

Medium
Category
Data Exfiltration
Confidence
60% 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 · scripts/weather.sh (reported line 14)May include surrounding context.

sh
DAYS="${4:-3}"
UNITS="${5:-fahrenheit}"

BASE="https://api.open-meteo.com/v1/forecast"

if [ "$UNITS" = "fahrenheit" ]; then
  TEMP_UNIT="&temperature_unit=fahrenheit&wind_speed_unit=mph&precipitation_unit=inch"

External Transmission

Medium
Category
Data Exfiltration
Confidence
60% 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 · scripts/weather_strip.py (reported line 92)May include surrounding context.

python
DAYS="${4:-3}"
UNITS="${5:-fahrenheit}"

BASE="https://api.open-meteo.com/v1/forecast"

if [ "$UNITS" = "fahrenheit" ]; then
  TEMP_UNIT="&temperature_unit=fahrenheit&wind_speed_unit=mph&precipitation_unit=inch"

Missing User Warnings

Low
Category
Not specified by scanner
Confidence
87% confidence
Finding

The script sends the supplied latitude and longitude to the Open-Meteo API via curl, but the only comments describe usage and units, not that location data is transmitted to a third-party service. There is no confirmation prompt, visible notice, or explicit warning comment about this outbound data transfer.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.