The Agent Skills Directory

[SAFE]: The skill acts as a style enforcer and linter for text generation. It defines a set of identity invariants and anti-patterns to guide the agent's output. These instructions are consistent with the skill's stated purpose and do not attempt to bypass security filters.
[SAFE]: The skill references a CLI tool, voice-scorer, to analyze document quality. This follows standard development patterns for automated linting and does not involve downloading remote code or executing unauthorized commands.
[SAFE]: Analysis of the indirect prompt injection attack surface shows that while the skill processes user-supplied documents, it uses them as intended for stylistic evaluation.
Ingestion points: Processes file paths of user documents (e.g., READMEs, design docs) in the SKILL.md usage examples.
Boundary markers: Not explicitly mentioned in the instructions, which is common for document-processing skills.
Capability inventory: Execution of the voice-scorer shell command.
Sanitization: Not specified, as the skill functions as an evaluator of existing text.
Assessment: The ingestion of documents is the primary functional requirement of the skill and does not present an elevated risk.

voice-enforcement