Composio Connect
Pass
Audited by Gen Agent Trust Hub on Nov 3, 2026
Risk Level: SAFEREMOTE_CODE_EXECUTIONEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTION
Full Analysis
- [REMOTE_CODE_EXECUTION]: The skill includes instructions to install the Composio CLI using a piped shell command:
curl -fsSL https://composio.dev/install | bash. While this pattern involves executing remote code, the source is the official domain for the tool provided by the vendor, making it a standard installation procedure for this skill. - [EXTERNAL_DOWNLOADS]: The skill downloads an installation script from
https://composio.dev/installto set up the execution environment. - [COMMAND_EXECUTION]: The skill utilizes several CLI commands (
composio login,composio tools search,composio tools execute) to perform its primary functions of authentication and tool interaction. - [INDIRECT_PROMPT_INJECTION]: The skill is designed to process data from external applications (Slack, GitHub, Gmail, etc.), which creates an attack surface for indirect prompt injection.
- Ingestion points: External data retrieved via
composio tools execute(SKILL.md). - Boundary markers: Absent; no specific delimiters or warnings to ignore instructions within retrieved data are provided.
- Capability inventory: The skill can execute CLI tools, bash scripts, and python scripts (SKILL.md).
- Sanitization: No explicit sanitization or filtering of external content is defined in the instructions.
- [DYNAMIC_EXECUTION]: Best practices in the skill suggest offloading data analysis to inline bash or python scripts, which involves the dynamic generation and execution of code based on processed data.
Audit Metadata