The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to interact with and process data that could originate from untrusted sources, creating a potential injection surface.
Ingestion points: Build logs (via kubectl logs), source code, and build contexts (via trivy and syft) are identified as data entry points.
Boundary markers: The provided examples do not include explicit delimiters or instructions to ignore embedded commands within the processed data.
Capability inventory: The skill utilizes several powerful tools including npm for builds, trivy for scanning, syft for SBOM generation, cosign for artifact signing, and kubectl for Kubernetes cluster management.
Sanitization: While the documentation recommends masking PII and secrets in logs, the execution examples do not show explicit sanitization or escaping of the ingested content.
[COMMAND_EXECUTION]: The skill provides numerous example commands for local execution to perform security checks and deployment verification.
The examples include the use of npm ci, trivy, syft, cosign, and various kubectl subcommands.
These are standard DevSecOps practices and the skill is described as documentation-only, with no bundled executable scripts.

openclaw-deployment-hardening