Tp2
- Category
- MCP Tool Poisoning
- Confidence
- 85% confidence
- Finding
Mixing characters from multiple Unicode scripts in a single identifier is a common technique to create visually ambiguous tool names.
- Content
Security audit
Security checks for vulnerabilities and agentic risk
ClawScan marks this skill suspicious because a diagnostic option may expose temporary upload credentials, while some scanner coverage is incomplete.
Consider the suspicious verdict before installing. In particular, avoid using `--print-curl` where its output could be saved or shared, since it may expose temporary credentials. The scan inspected all three listed components, but unresolved references made the analysis partial, so the evidence is incomplete.
Mixing characters from multiple Unicode scripts in a single identifier is a common technique to create visually ambiguous tool names.
Mixing characters from multiple Unicode scripts in a single identifier is a common technique to create visually ambiguous tool names.
Mixing characters from multiple Unicode scripts in a single identifier is a common technique to create visually ambiguous tool names.
Mixing characters from multiple Unicode scripts in a single identifier is a common technique to create visually ambiguous tool names.
Mixing characters from multiple Unicode scripts in a single identifier is a common technique to create visually ambiguous tool names.
Mixing characters from multiple Unicode scripts in a single identifier is a common technique to create visually ambiguous tool names.
Mixing characters from multiple Unicode scripts in a single identifier is a common technique to create visually ambiguous tool names.
Without declared permissions the skill's intent is opaque and cannot be validated.
The --print-curl branch prints the complete command, including the temporary security token and signed Authorization header, despite the documented promise not to echo the token. If captured in agent transcripts, terminal logs, or support reports, these values can permit replay of the signed upload while its signature and credentials remain valid; the SecretKey itself is not printed.
subprocess module calls execute external commands. Without careful input validation, this enables command injection.
return
try:
out = subprocess.run(curl_cmd, capture_output=True, text=True, timeout=120)
except subprocess.TimeoutExpired:
fail("上传超时(120s)。检查网络后重试。")