The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes multiple local Python scripts and shell commands (e.g., uv run python test_memory.py, bash deploy.sh) to test features and deploy AWS resources. This includes interaction with the agentcore CLI tool.\n- [EXTERNAL_DOWNLOADS]: The skill uses uv run, which automatically manages Python environments and downloads necessary package dependencies from official registries like PyPI during the workshop execution.\n- [DATA_EXFILTRATION]: The skill reads local configuration files such as .bedrock_agentcore.yaml and requests a user-provided SES sender email address. These actions are intended for configuring and testing the workshop's cloud infrastructure.\n- [PROMPT_INJECTION]: The skill accepts user-supplied $ARGUMENTS to control which workshop steps are run. It includes logic for the agent to parse, expand, and validate these inputs, which helps ensure they are treated as configuration parameters rather than executable commands.

run-workshop