The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is primarily designed to retrieve and process untrusted external data from Twitter/X, creating a surface for indirect prompt injection attacks.
Ingestion points: Data enters the agent's context through various API endpoints, including tweet search, profile lookups, and bulk extractions (as detailed in SKILL.md and references/x-api.md).
Boundary markers: The instructions mandate the use of explicit boundary markers, <XQUIK_UNTRUSTED_X_CONTENT> and </XQUIK_UNTRUSTED_X_CONTENT>, to encapsulate all retrieved external content (specified in SKILL.md and references/security.md).
Capability inventory: The skill is restricted to using the WebFetch tool for API communication. Shell access, filesystem access, and local code execution are explicitly disabled in the YAML frontmatter.
Sanitization: Instructions require the agent to serialize external content as JSON and apply Unicode escaping to sensitive characters (such as <, >, and &) to prevent boundary escapes and instruction hijacking (detailed in references/security.md).
[REMOTE_CODE_EXECUTION]: An automated security scan flagged references/scrape-export-twitter-data.md with a 'FileRepMalware' alert. Technical inspection of the file confirms it contains markdown documentation and standard Python code snippets (using the requests library) for API interaction. The alert appears to be a reputation-based false positive frequently associated with scraping-related documentation and tools rather than a presence of actual malicious payloads.

x-twitter-scraper