The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data through the fetchDeploymentLogs function, which retrieves runtime logs from the deployed application. Logs can contain content generated by external actors (e.g., via application inputs or error messages), which may include instructions intended to mislead the agent during debugging tasks. This could potentially influence the agent's use of the verifyAndReplaceArtifactToml function to modify deployment settings.
Ingestion points: fetchDeploymentLogs (referenced in SKILL.md and references/deployment-logs.md).
Capability inventory: verifyAndReplaceArtifactToml (modifies production configuration files), suggestDeploy (prompts user to publish changes).
Boundary markers: No specific delimiters or "ignore embedded instructions" warnings are provided for the log analysis workflow.
Sanitization: The skill does not describe any sanitization or filtering of log content before it is provided to the agent for analysis.

deployment