The Agent Skills Directory

[COMMAND_EXECUTION]: The skill employs standard shell commands including mkdir, rsync, cp, and python3 to set up environment templates and execute local scripts. These operations are scoped to the project directories and plugin root.
[INDIRECT_PROMPT_INJECTION]: The skill ingests data from external FOCUS-formatted CSV files (finops/focus/*.csv), specifically parsing a Tags column which could contain arbitrary content provided by cloud providers or resource owners.
Ingestion points: Data is read from finops/focus/*.csv using csv.DictReader in Step 3.
Boundary markers: No explicit boundary markers or instructions to ignore embedded content are defined in the summary script.
Capability inventory: The skill has the ability to execute shell commands, read/write files in the local repository, and generate documentation pages.
Sanitization: The script uses json.loads() for parsing tag data, but does not sanitize the resulting strings before they are printed for the agent's context.
[DYNAMIC_EXECUTION]: The skill generates and executes a transient Python script using a shell HEREDOC (`python3
... <<'EOF'`) to summarize costs. The script is static, defined entirely within the skill instructions, and is used for data processing rather than arbitrary code execution.

showback