The Agent Skills Directory

[SAFE]: The skill is a standard utility focused on text processing tasks such as fixing typos, removing double spaces, and standardizing punctuation. It includes specific instructions to preserve technical terms and the author's original voice.
[COMMAND_EXECUTION]: The skill's workflow utilizes the grep command-line tool to identify formatting issues (specifically em dashes) within local files. This is a standard and expected use of local utilities for this type of task.
[PROMPT_INJECTION]: The skill processes untrusted data from user-provided files, which represents a surface for indirect prompt injection.
Ingestion points: Reads content from .md and .txt files.
Boundary markers: The skill does not implement specific delimiters or warnings to ignore instructions embedded within the target files.
Capability inventory: The skill can read file contents, search using grep, and perform in-place edits to files.
Sanitization: No sanitization is performed on the input text before it is processed by the agent.
[DATA_EXFILTRATION]: No network operations, sensitive file access, or credential exposure were detected. The skill's scope is strictly limited to local text and markdown files.

prose-fix