The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests data from local files such as tests/judgments.csv and tests/baseline.json to populate search queries and ratings for the Rank Evaluation API. While this creates an attack surface for indirect prompt injection or SSRF if the local files are maliciously crafted, the current implementation is focused on local development and testing workflows against user-controlled search clusters (e.g., localhost:9200 or localhost:8983). No sanitization is present, which is typical for internal developer tooling.
[COMMAND_EXECUTION]: The skill provides Python code snippets and shell commands (e.g., bin/solr start, curl -XPUT) intended for the developer to execute against their local search environment. These are standard operations for search engine configuration and testing.
[EXTERNAL_DOWNLOADS]: The skill references and links to official documentation for Elasticsearch, OpenSearch, and Apache Solr, as well as well-known search engineering tools like Quepid and Splainer on GitHub. These references are informative and point to trusted or well-known technology services.

elasticsearch-relevance-tests