The Agent Skills Directory

[SAFE]: The skill consists entirely of Markdown documentation and YAML configuration. No executable scripts (.py, .js, .sh) or binaries are included.- [PROMPT_INJECTION]: No malicious instructional overrides or jailbreak attempts were found. Indirect injection surface analysis: 1. Ingestion points: user manuscript notes and drafts (referenced in SKILL.md); 2. Boundary markers: absent; 3. Capability inventory: none (no subprocess, network, or file access found in any of the 42 files); 4. Sanitization: none mentioned. The complete lack of dangerous capabilities renders the ingestion surface safe.- [DATA_EXFILTRATION]: No network access or data exfiltration patterns were detected. All external URLs (such as those to Notion and GitHub) lead to legitimate academic writing resources and documentation.- [REMOTE_CODE_EXECUTION]: There are no remote code downloads, package installations, or execution patterns within the skill files.

nature-writing