The Agent Skills Directory

[SAFE]: The skill implements sound security practices for secret management. It instructs the agent to avoid passing API keys via command-line flags, recommending environment variables or the ~/.config/dune/config.yaml file instead. It also provides instructions to redact sensitive tokens from any output presented to the user.
[SAFE]: All external references and documentation links point to official Dune Analytics domains (dune.com, docs.dune.com), which are consistent with the skill's stated authorship and purpose.
[SAFE]: The skill processes external blockchain data, which constitutes an indirect prompt injection surface. However, this is necessary for its function as a data-retrieval tool and does not indicate malicious intent.
Ingestion points: Query results retrieved from Dune via dune query run, dune query run-sql, and dune execution results in the query-execution.md reference file.
Boundary markers: No explicit markers are used to delimit data from instructions in the prompt.
Capability inventory: The skill has access to the Bash tool to execute dune and curl commands as defined in the SKILL.md frontmatter.
Sanitization: The skill does not specify sanitization or filtering of query results before processing.

dune