The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to process and format legal documents and case information which may originate from untrusted external sources (e.g., drafts from a 'Schreib-Canvas').
Ingestion points: Processes document content and case details provided by the user in SKILL.md.
Boundary markers: There are no explicit delimiters or instructions to ignore embedded commands within the processed document text.
Capability inventory: The skill invokes several other internal skills (e.g., kanzlei-allgemein-schriftsatz-turbo, kanzlei-allgemein-bea-journal) and writes structured data to assets/templates/output-versandprotokoll.md.
Sanitization: The skill does not mention specific sanitization or filtering steps for the text it processes before passing it to other tools.
[SAFE]: The skill includes positive security practices by explicitly warning users not to input sensitive credentials such as beA software tokens, certificates, PINs, or passwords into the chat interface.
[SAFE]: All referenced tools and external skills (kanzlei-allgemein-*) appear to be part of the vendor's internal ecosystem for legal automation and represent standard functionality.

output-versand