The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from Amazon search results and product details, creating an attack surface where malicious content in book listings could influence agent behavior.
Ingestion points: Data enters the context via the JSON output files of bsr-scraper.py and keyword-volume.py (e.g., /tmp/niche-scout-bsr.json).
Boundary markers: The instructions lack explicit boundary markers or warnings to the agent to treat the external data as untrusted.
Capability inventory: The skill uses Python scripts with network access via the requests library and file write capabilities to the /tmp/ directory.
Sanitization: While bsr-scraper.py performs basic normalization on field names to handle Unicode artifacts (like Right-to-Left Marks), it does not sanitize the values (titles, descriptions, keywords) retrieved from Amazon search results before they are presented to the LLM.
[EXTERNAL_DOWNLOADS]: The skill relies on an external, third-party dependency: the curious_coder~amazon-scraper actor on Apify. This community-contributed scraping tool introduces a dependency on an unverified third party whose output directly influences analysis results and agent decision-making.

niche-scout