The Agent Skills Directory

[CREDENTIALS_UNSAFE]: The script scripts/env-loader.cjs is designed to load environment variables from multiple paths, including $HOME/.claude/.env (the global configuration file for the agent platform) and skill-specific .env files. Accessing global configuration files to load secrets increases the risk of exposing sensitive platform-level credentials.
[COMMAND_EXECUTION]: The script scripts/gsc-content-analyzer.cjs utilizes dynamic module loading by modifying the Node.js module resolution path (module.paths.unshift) and using require() on computed paths to load libraries and scripts from a sibling skill directory (seo-optimization). This practice executes code from unverified external sources at runtime and bypasses standard dependency controls.
[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection within scripts/gsc-content-analyzer.cjs.
Ingestion points: The agent context is populated with data from Google Search Console API responses (e.g., search queries and page titles).
Boundary markers: The ingested data is placed directly into Markdown tables in the generated report without explicit delimiters or instructions to ignore embedded commands.
Capability inventory: The script possesses file-writing capabilities (fs.writeFileSync), which could be leveraged to persist malicious data or instructions.
Sanitization: There is no evidence of validation or escaping for the external content before it is interpolated into the reports.
[NO_CODE]: The SKILL.md file provides examples for scripts/content-planner.py and scripts/content-audit.py, but these files were not included in the provided skill package, preventing a complete security audit of the content planning functionality.

content-marketing