The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes content from Markdown files provided by the user, creating a surface for indirect prompt injection. Malicious instructions embedded in these files could attempt to manipulate the generated jargon report.
Ingestion points: Markdown files identified via Glob and accessed through the Read tool in SKILL.md (Step 1).
Capability inventory: The skill is restricted to file system reading and searching (Read, Grep, Glob). It does not possess capabilities for command execution, file writing, or network exfiltration.
Boundary markers: The skill does not employ explicit delimiters to separate user data from instructions, although it does define logical exclusions for code blocks and specific metadata contexts (Step 2.1).
Sanitization: No sanitization or filtering is applied to the ingested text before it is analyzed by the agent.
[EXTERNAL_DOWNLOADS]: The skill references external style guide resources from the vendor's official documentation site.
Evidence: Source URLs pointing to elastic.co documentation are included in the skill's metadata.

docs-flag-jargon-skill