The Agent Skills Directory

[PROMPT_INJECTION]: The skill possesses an indirect prompt injection surface because it is designed to ingest and process untrusted data from external sources such as articles, books, and research papers. Maliciously crafted instructions within these sources could potentially influence the agent's extraction and categorization logic during the enrichment process. * Ingestion points: Article/Book processing and Research Integration workflows (SKILL.md, references/enrichment-workflow.md). * Boundary markers: None identified. The instructions do not employ specific delimiters or 'ignore' instructions for external content. * Capability inventory: The agent is instructed to use tools like qmd and obsidian and performs file modifications (write/update/delete) on the local filesystem. * Sanitization: None identified. The skill does not describe processes for validating or cleaning external data before extraction.
[COMMAND_EXECUTION]: The skill's maintenance and validation procedures involve the execution of local shell commands and Python scripts. * Evidence: The documentation specifies using qmd search, qmd query, and obsidian backlinks for search and deduplication operations, and python3 to run find_broken_links.py, find_orphans.py, and check_frontmatter.py for vault audits (references/duplicate-detection.md, references/enrichment-workflow.md).

ideaverse-enrichment