pp-airbyte-admin

Pass

Audited by Gen Agent Trust Hub on Aug 20, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONDATA_EXFILTRATION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill instructs the agent to download and install the CLI and MCP server from external sources controlled by the author.
  • Fetches the CLI via npx -y @mvanhorn/printing-press-library.
  • Installs Go modules from github.com/mvanhorn/printing-press-library/library/cloud/airbyte-admin/cmd/airbyte-admin-pp-cli and github.com/mvanhorn/printing-press-library/library/cloud/airbyte-admin/cmd/airbyte-admin-pp-mcp.
  • [REMOTE_CODE_EXECUTION]: The skill facilitates the installation and execution of remote code through standard package managers.
  • Uses npx to execute code directly from the NPM registry.
  • Uses go install to compile and install binaries from a remote Git repository.
  • [DATA_EXFILTRATION]: The CLI tool includes a built-in feature to transmit command output to arbitrary external servers.
  • The --deliver webhook:<url> flag allows the agent to POST the results of Airbyte inspection (which may include sensitive workspace, user, and connection metadata) to a user-specified URL.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes data from external Airbyte APIs, which could serve as a vector for indirect prompt injection.
  • Ingestion points: Reads metadata from Airbyte workspaces, sources, destinations, and connections (SKILL.md).
  • Boundary markers: None explicitly mentioned in the instructions for sanitizing Airbyte API responses.
  • Capability inventory: Execution of shell commands via airbyte-admin-pp-cli and network operations via the webhook sink.
  • Sanitization: None described for external content interpolated into agent context.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 20, 2026, 08:25 AM
Security Audit — agent-trust-hub — pp-airbyte-admin