The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill recommends installing a supplemental vendor resource using the command npx skills add kestra-io/agent-skills@migrate-kestra-2. This targets the kestra-io organization's repository, which is consistent with the skill's authorship.
[COMMAND_EXECUTION]: The skill instructions drive the use of the kestractl CLI tool, involving the construction of shell commands that incorporate user-provided variables such as flow IDs, namespaces, and file paths. Standard operational patterns like kestractl flows deploy and kestractl nsfiles upload are defined.
[INDIRECT_PROMPT_INJECTION]: The skill interacts with external Kestra environments, creating an attack surface where data retrieved from those environments could potentially influence agent behavior. The skill mitigates this through operational guardrails such as manual confirmation for production actions and the use of validation commands (flows validate).
Ingestion points: Data returned from Kestra API through kestractl commands (flows, execution logs, namespace files).
Boundary markers: The skill encourages the use of the --output json flag for automation, which helps in structured data handling.
Capability inventory: Shell command execution via kestractl and skill installation via npx.
Sanitization: No explicit string sanitization for shell arguments is defined in the instructions.

kestra-ops