The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes the mthds-agent CLI to perform several operations. These include fetching input schemas with mthds-agent pipelex inputs and executing pipelines using mthds-agent pipelex run pipe. The commands are used to process local bundles and manage workflow execution.
[PROMPT_INJECTION]: The skill possesses a surface for indirect prompt injection as it processes untrusted data provided by users.
Ingestion points: Data is ingested through the inputs.json file located in the bundle directory or via the --inputs JSON flag in the command line.
Boundary markers: No specific boundary markers or 'ignore' instructions are defined within the skill to isolate untrusted content from the pipeline execution context.
Capability inventory: The skill triggers shell command execution through the mthds-agent tool and performs file system operations (writing live_run.json and .html files).
Sanitization: The instructions do not describe any sanitization or validation of the input data before it is passed to the execution tool.

run