The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to execute local scripts located in ~/.claude/skills/econstack/bin/ silently in the background for version checking and project state management.
[REMOTE_CODE_EXECUTION]: The preamble includes the command eval "$(~/.claude/skills/econstack/bin/econstack-slug)". This evaluates the output of a local binary as shell code, which is a dynamic execution pattern that could lead to unauthorized command execution if the output is manipulated.
[PROMPT_INJECTION]: The skill ingests untrusted data from web searches and external R packages (ONS, FRED, Comtrade) to generate briefing notes. This creates an indirect prompt injection surface where malicious data could attempt to influence agent behavior.
Ingestion points: WebSearch and WebFetch tools, and external R package data specified in Step 3 of SKILL.md.
Boundary markers: No explicit delimiters or instructions are provided to the agent to disregard potential embedded instructions within the fetched data.
Capability inventory: The skill allows access to Bash, Read, and Write tools throughout its execution.
Sanitization: No sanitization or validation logic is specified for the external content before it is incorporated into the output files.

briefing-note