The Agent Skills Directory

[SAFE]: The skill's functionality is consistent with its stated purpose of implementing and auditing JIT access controls. The Python scripts (scripts/agent.py, scripts/process.py) perform legitimate data analysis and lifecycle management tasks without suspicious network activity, credential harvesting, or unauthorized privilege escalation.
[INDIRECT_PROMPT_INJECTION]: The script scripts/agent.py ingests data from external JSON files to perform its audit functions, which represents a potential attack surface for indirect prompt injection if the output were processed by an LLM without safeguards.
Ingestion points: Data is loaded from files specified by the --requests and --privileges command-line arguments in scripts/agent.py.
Boundary markers: Not applicable, as this is a standalone analytical script.
Capability inventory: File system access is limited to reading input logs and writing a summary report via json.dump; no network or subprocess capabilities are utilized by the audit script.
Sanitization: The script does not sanitize data read from the JSON files before including it in the generated report, though the risk is negligible in this context.

implementing-just-in-time-access-provisioning