The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses the jq utility to extract package names from package.json and grep to find instances of specific libraries in the source code. These are standard, read-only analytical operations within a development environment.
[EXTERNAL_DOWNLOADS]: References well-known and reputable industry resources including State of JS, State of CSS, and the Thoughtworks Technology Radar. These sources are used exclusively for fetching statistical and advisory data to inform the technology audit.
[PROMPT_INJECTION]: The skill processes information from local project configuration files and external survey results. While this constitutes an indirect data ingestion surface, the instructions target highly reputable community sources and do not include capabilities that would allow for privilege escalation or malicious command execution based on that data.
Ingestion points: Project package.json file and JSON results from the Devographics API (assets.devographics.com).
Boundary markers: None specified for the data ingestion, which is common for descriptive/analytical tasks.
Capability inventory: Limited to local read operations and shell-based string processing (jq, grep, sort).
Sanitization: Not explicitly defined, though the scope of operations is inherently restricted to text analysis and documentation generation.

tech-trend-watch