microsoft-graph-api
Pass
Audited by Gen Agent Trust Hub on Oct 11, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill instructs the installation of the
@membranehq/clipackage from the NPM registry. This is the official command-line interface for the Membrane platform, which is the vendor of this skill. - [COMMAND_EXECUTION]: The skill relies on the execution of shell commands through the
membraneCLI to manage authentication (membrane login), handle connections (membrane connection ensure), and execute tasks (membrane action run). - [INDIRECT_PROMPT_INJECTION]: The skill has a surface for indirect prompt injection as it interpolates user-provided data into shell commands.
- Ingestion points: User-provided search intents in
membrane action list, JSON payload inputs inmembrane action run, and endpoint paths inmembrane request(found in SKILL.md). - Boundary markers: No explicit boundary markers or instructions are provided to the agent to treat external inputs as untrusted data.
- Capability inventory: The skill allows the agent to execute actions and make arbitrary HTTP requests via the Membrane proxy, providing a bridge between local command execution and remote API interaction.
- Sanitization: There are no documented sanitization or validation steps for the inputs before they are interpolated into the CLI commands.
- Ingestion points: User-provided search intents in
Audit Metadata