seo-seranking

Warn

Audited by Gen Agent Trust Hub on Sep 10, 2026

Risk Level: MEDIUMCOMMAND_EXECUTIONDYNAMIC_EXECUTIONDATA_EXFILTRATIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [DATA_EXFILTRATION]: The skill instructions require access to a sensitive file path ~/.claude/settings.json to verify or retrieve the SERANKING_API_KEY stored under environment variables.
  • [COMMAND_EXECUTION]: The documentation directs the agent or user to execute installation scripts extensions/seranking/install.sh or install.ps1 for initial environment configuration.
  • [DYNAMIC_EXECUTION]: The skill uses the command claude-seo run dataforseo_costs.py to execute a local script for tracking API usage costs.
  • [INDIRECT_PROMPT_INJECTION]: The skill accepts untrusted user inputs such as brands, keywords, and URLs which are interpolated into commands for external API interaction.
  • Ingestion points: User data is ingested via the <brand>, <keyword>, and <url> parameters defined in the routing commands within SKILL.md.
  • Boundary markers: There are no defined delimiters or instructions to treat these inputs strictly as data, potentially allowing embedded instructions to be processed by the agent.
  • Capability inventory: The skill has the capability to execute local scripts and read sensitive configuration files on the host system.
  • Sanitization: No evidence of input validation, filtering, or sanitization is present in the skill instructions.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Sep 10, 2026, 09:08 PM
Security Audit — agent-trust-hub — seo-seranking