pdfco
Pass
Audited by Gen Agent Trust Hub on May 5, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONREMOTE_CODE_EXECUTIONPROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill guides the user to install '@membranehq/cli' from the public NPM registry. This is an official command-line tool provided by the vendor (membranedev) to interact with their integration platform.
- [COMMAND_EXECUTION]: Shell commands are used to manage the full lifecycle of the integration, including authentication ('membrane login'), connecting to services ('membrane connect'), and executing specific PDF tasks ('membrane action run').
- [REMOTE_CODE_EXECUTION]: The skill utilizes the 'membrane action create' feature, which generates integration code on the Membrane platform to be executed through the CLI. This is a standard and intended functionality of the Membrane ecosystem for creating custom API connectors.
- [PROMPT_INJECTION]: The skill possesses an indirect prompt injection surface as it is designed to ingest and process data from external PDF files. 1. Ingestion points: External data processed via PDF.co extraction, conversion, and information retrieval actions (SKILL.md). 2. Boundary markers: None identified in the provided instructions; data is processed as-is. 3. Capability inventory: Execution of shell commands and network interaction via the Membrane CLI (SKILL.md). 4. Sanitization: No explicit sanitization or validation of the extracted PDF content is described in the skill instructions.
Audit Metadata