The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill exhibits an attack surface for indirect prompt injection as it processes untrusted external data.
Ingestion points: Reads project context from local files (seo-project-context.md) and analyzes tenant repository contents using tools like rg (ripgrep).
Boundary markers: The skill includes instructions to prioritize repository facts over guesses and warns against inventing patterns, which serves as a basic behavioral boundary.
Capability inventory: The skill has the capability to execute shell commands via npx, write files to the repository (e.g., seo-blog.config.json, workflows), and perform network operations via the vendor CLI.
Sanitization: No explicit sanitization or validation logic for the input context or repository data is described in the prompt instructions.
[EXTERNAL_DOWNLOADS]: The skill fetches resources from external sources associated with the vendor.
It installs and updates itself from a GitHub repository (Wotaso/seodrafts-agent-skills).
It resolves and executes the latest version of the @wotaso/seo-blog-admin-cli package from the npm registry using npx.
[COMMAND_EXECUTION]: The skill relies on executing various shell commands to perform its tasks.
It uses npx to run CLI tools for project creation, token rotation, and publishing initialization.
It uses rg (ripgrep) to discover repository patterns and configurations.

seo-blog-admin-projects