The Agent Skills Directory

[COMMAND_EXECUTION]: The skill is configured with the Bash tool, allowing the agent to execute shell commands to manage data pipelines and schedules. This is a powerful capability that requires user oversight.
[CREDENTIALS_UNSAFE]: In Step 0 (Detect Environment), the skill instructs the agent to "Check for data warehouse configs: BigQuery, Redshift, Snowflake connection details." This directive encourages the agent to search for and potentially load sensitive credentials and connection strings into the model context.
[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection (Category 8) because it is designed to ingest and process data from external, potentially untrusted sources.
Ingestion points: Reads data from APIs, databases, and files during the 'Extract' phase defined in SKILL.md.
Boundary markers: The instructions do not define clear delimiters or warning markers to prevent the agent from obeying instructions embedded within the extracted data.
Capability inventory: The skill uses Read, Write, Edit, Bash, and WebFetch tools as defined in the SKILL.md frontmatter.
Sanitization: There are no explicit instructions for the agent to sanitize or validate the content extracted from external sources before processing it.

flux-pipeline