The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill extracts content from arbitrary external websites (competitor landing pages, blogs, and pricing pages) using the actionbook tool. This content is then processed by the agent to generate competitive analysis reports. This creates a surface for indirect prompt injection, where a malicious site could host hidden instructions designed to influence the agent's behavior or output.
Ingestion Points: The skill uses actionbook extract <url> in SKILL.md to fetch data from user-provided URLs.
Boundary Markers: There are no explicit instructions or delimiters defined to separate the untrusted external content from the agent's instructions during synthesis.
Capability Inventory: The agent is authorized to execute shell commands (actionbook) and generate synthesized reports based on external data.
Sanitization: No sanitization, filtering, or validation of the scraped web content is described.
[EXTERNAL_DOWNLOADS]: The skill instructs the user to install a third-party command-line interface tool, @actionbookdev/cli, via the NPM registry (npm install -g @actionbookdev/cli).
[COMMAND_EXECUTION]: The core functionality of the skill relies on executing shell commands via the actionbook CLI to perform web extraction and search tasks.

gr-competitor