Lp1
High
- Category
- MCP Least Privilege
- Confidence
- 75% confidence
- Finding
The skill uses 'env' capability that is not listed in its permissions. This may indicate deceptive intent or missing permission declarations.
- Content
Security audit
KnowAir Weather 明气天气
Security checks for vulnerabilities and agentic risk
Overview
This skill appears to be a straightforward air-quality lookup tool that sends coordinates and a Caiyun API token to Caiyun services as disclosed.
Install only if you are comfortable providing a Caiyun API token and having your requested coordinates sent to Caiyun. Prefer CAIYUN_TOKEN if you do not want the skill reading the optional ~/.config/knowair/token file, and treat the exercise recommendations as air-quality guidance rather than medical advice.
Vulnerability Patterns
- 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
- Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs
Vulnerability Patterns
- Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
- Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
- Rogue AgentSelf-Modification, Session Persistence
- Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
- MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
Findings (6)
Lp1
High
- Category
- MCP Least Privilege
- Confidence
- 75% confidence
- Finding
The skill uses 'network' capability that is not listed in its permissions. This may indicate deceptive intent or missing permission declarations.
- Content
Context-Inappropriate Capability
Medium
- Category
- Not specified by scanner
- Confidence
- 88% confidence
- Finding
The manifest describes an air-quality forecast skill and allows only Bash(python3:*) and Read, while the documentation says the token may be stored in ~/.config/knowair/token. Accessing a specific local credential file is an additional capability beyond the obvious need to query air-quality data, especially since the manifest's primary credential path is the CAIYUN_TOKEN environment variable.
- Content
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
python import urllib.request import urllib.error API_BASE = "https://api.caiyunapp.com/v2.6" STATION_API = "https://singer.caiyunhub.com/v3/aqi/forecast/station"
Session Persistence
Medium
- Category
- Rogue Agent
- Confidence
- 60% confidence
- Finding
Skill establishes unauthorized persistence across sessions via cron jobs, startup scripts, or state files. Session persistence allows an attacker to maintain access beyond the current interaction.
- Content
python token = get_token() if not token: print(json.dumps({"error": "No API token found. Set CAIYUN_TOKEN env var or create ~/.config/knowair/token"})) sys.exit(2) hours = max(1, min(360, args.hours))
Vague Triggers
Low
- Category
- Not specified by scanner
- Confidence
- 90% confidence
- Finding
The description says to use the skill when the user asks 'is it safe to exercise outdoors,' which is broader than a clearly scoped air-quality invocation and could overlap with general health, weather, or activity advice. The trigger list is example-based but does not provide exclusion conditions or boundaries for when the skill should not activate.
- Content
Static analysis
No suspicious patterns detected.