Lp3
Medium
- Category
- MCP Least Privilege
- Confidence
- 90% confidence
- Finding
- The skill documentation explicitly directs the agent to invoke a shell script that performs outbound HTTP requests with curl, yet no declared permissions are present to reflect those capabilities. This creates a transparency and policy-enforcement gap: a host system or reviewer may underestimate the skill's ability to execute commands and access the network, increasing the chance of unintended token-backed API actions.
