The Agent Skills Directory

[SAFE]: No security vulnerabilities, malicious patterns, or unauthorized activities were detected. The skill is composed of markdown-based instructions for styling and auditing text.
[PROMPT_INJECTION]: The skill identifies a potential indirect prompt injection surface as it is designed to ingest and process user-provided drafts.
Ingestion points: Workflow 1 in SKILL.md ('Ingestion & Analysis') processes external text or target files.
Boundary markers: Absent; there are no specific instructions in the provided files for the agent to use protective delimiters around user input.
Capability inventory: The skill is restricted to text processing within the LLM; no executable scripts, system calls, or network operations are included.
Sanitization: The references/audit_checklist.md provides a 'Semantic Preservation Audit' which requires the agent to verify that the output remains factually consistent with the original text, providing a safeguard against instruction-based hijacking.
[EXTERNAL_DOWNLOADS]: The documentation provides installation instructions via a GitHub repository (github.com/kshanxs/skill-dojo). This reference targets the author's official repository on a well-known service.

calligraph-sensei