The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses shell commands to execute its various scraping modules.
Evidence: SKILL.md contains multiple python3 calls (Phases 1, 2, and 3) to run scripts located in skills/ subdirectories.
[EXTERNAL_DOWNLOADS]: The skill is designed to fetch data from external sources including competitor blogs, LinkedIn, and Twitter/X.
Evidence: The configuration in Phase 0 and the commands in Phases 1-3 target external URLs and handles for scraping via the Apify API.
[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it ingests untrusted content from the web and processes it to create a synthesized summary.
Ingestion points: Data is pulled from external URLs (blogs) and social media profiles in Phase 1, Phase 2, and Phase 3.
Boundary markers: Absent. The instructions in Phase 4 (Analyze & Synthesize) do not specify the use of delimiters or provide instructions for the agent to ignore potentially malicious commands embedded in the scraped text.
Capability inventory: The skill allows for local shell command execution (python3 scripts) and file writes to the clients/ directory.
Sanitization: Absent. There is no mention of sanitizing, escaping, or validating the ingested content before it is passed to the LLM for synthesis.

competitor-content-tracker