The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill fetches keyword data from a remote API defined by the NEXSCOPE_PROXY_BASE environment variable.
[DATA_EXFILTRATION]: While the skill communicates with an external API, it is restricted to search parameters and marketplace identifiers necessary for its primary function. It uses the NEXSCOPE_API_KEY provided via environment variables to authenticate these requests.
[PROMPT_INJECTION]: The skill processes external data from the Jungle Scout API and interpolates keyword names directly into a Markdown report for the agent's context. This creates a surface for indirect prompt injection if the API source were to return malicious instructions in the keyword name fields.
Ingestion points: scripts/keyword_research.py fetches data via the NexScope proxy endpoint.
Boundary markers: The output uses Markdown tables and headers to structure data, but lacks explicit boundary markers to isolate potentially untrusted keyword strings.
Capability inventory: The agent is granted Bash, Read, and Write tools as specified in SKILL.md.
Sanitization: There is no evidence of string sanitization or escaping for the name field returned by the API before it is rendered in the final report.

keyword-research