The Agent Skills Directory

[COMMAND_EXECUTION]: The skill implements dynamic context injection via !command syntax in SKILL.md. This script executes shell commands automatically when the skill is loaded to detect FUNDA_API_KEY by checking the environment and searching for .env files in the working directory or git common directory using git, grep, and pwd.
[EXTERNAL_DOWNLOADS]: The skill performs numerous remote requests using curl to api.funda.ai to fetch various financial datasets, news summaries, and social media content.
[DATA_EXFILTRATION]: Authentication tokens (API keys) are transmitted to the external domain https://api.funda.ai. Additionally, the skill includes a proxy configuration for the Claude API (/v1/claude/v1/messages) that routes LLM traffic through the vendor's infrastructure.
[COMMAND_EXECUTION]: The skill generates shell command strings for the agent to execute, primarily using curl piped into python3 -m json.tool for data processing and formatting.

funda-data