The Agent Skills Directory

[COMMAND_EXECUTION]: The skill constructs and executes shell commands that incorporate user-provided variables without specifying sanitization or escaping.
The command python scripts/dfs_client.py serp --keyword "<target keyword>" --limit 10 uses a keyword either supplied by the user or inferred from untrusted content. If the user provides a keyword containing shell metacharacters (e.g., "; command #), it could lead to arbitrary command execution.
The command python scripts/report_publish.py <report>.md uses a filename containing a <slug> derived from the user's content, which presents a similar injection risk if the slug is not sanitized.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process untrusted external data, creating an attack surface for indirect prompt injection.
Ingestion points: User-provided 'live URLs' and 'pasted drafts' in the SKILL.md instructions.
Boundary markers: None. The skill does not instruct the agent to use delimiters or ignore instructions embedded within the content being reviewed.
Capability inventory: The skill has the ability to execute local Python scripts (dfs_client.py, report_publish.py), perform network fetches via webfetch, and write files to the local file system.
Sanitization: None. There are no instructions to validate, filter, or escape the user-supplied content before it is used to generate command-line arguments or filenames.

content-review