The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes output from external command executions, creating a surface for untrusted data to enter the agent context.
Ingestion points: Data from mcloud deployments list, get, and build-logs as described in SKILL.md.
Boundary markers: The prompt suggests using JSON and jq for structured data but lacks specific delimiters or 'ignore' instructions for the data content.
Capability inventory: The agent uses Bash tools for mcloud and jq operations.
Sanitization: No sanitization or validation logic is specified for the ingested command output.
[COMMAND_EXECUTION]: The skill leverages specific CLI tools (mcloud and jq) to perform its tasks. These tools are scoped within the skill's defined context for deployment management.

mcloud-deployments