The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to use shell commands to facilitate documentation lookup and validation. Additionally, the examples/ext-exec.js file demonstrates the use of the xk6-exec extension, which allows load test scripts to execute arbitrary shell commands. While the provided examples use benign commands such as date and echo, this capability allows for system-level interaction.
[EXTERNAL_DOWNLOADS]: The SETUP.md file contains instructions to download the xk6 build tool and a documentation extension from official repositories within the Grafana GitHub organization. These downloads are performed from trusted infrastructure to enable core skill functionalities like documentation lookup.
[INDIRECT_PROMPT_INJECTION]: The examples/ext-crawler.js file provides a template for using xk6-crawler to retrieve content from external URLs, creating a surface where untrusted data could be ingested into the agent's context.
Ingestion points: The crawler.crawl(url) method in examples/ext-crawler.js and various http.get() calls in other examples fetch data from external sources.
Boundary markers: No specific delimiters or instructions to ignore embedded commands are present in the scripts that process external data.
Capability inventory: The skill's environment allows for shell command execution (via xk6-exec), file system writes (via agent tools used in Step 4), and network operations.
Sanitization: Examples show minimal sanitization, such as trimming whitespace from crawled titles, but do not demonstrate robust filtering of external content.

k6