The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes a local Python script (emulo.py) to perform preflight checks, prepare mining plans, and activate generated profiles. These commands are executed via the host's Python interpreter.
[DYNAMIC_EXECUTION]: The skill dynamically resolves the path for emulo.py relative to its own installation directory. It utilizes several subcommands (plugin validate-report, plugin cache-report, plugin validate-pack) to manage the workflow.
[INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from local .jsonl coding session history files. This presents a vulnerability surface for indirect prompt injection.
Ingestion points: Reads .jsonl session history files as source material.
Boundary markers: Uses instructions defined in MINING_PROMPT.md and reducer contracts to guide model processing.
Capability inventory: Execution of Python subprocesses for validation, caching, and activation.
Sanitization: Implements mandatory validation steps (validate-report, validate-pack) and requires an approval_hash to ensure the user has reviewed and authorized the execution plan before model work begins.

mine