sidework-rules-check
Pass
Audited by Gen Agent Trust Hub on Oct 4, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: No malicious patterns or security risks were identified. The skill follows best practices by using bundled scripts for data validation rather than relying solely on LLM reasoning.\n- [COMMAND_EXECUTION]: The skill executes a bundled Python script (
check_sidework_rules.py) to process user data. This execution is confined to the agent's local environment and uses predefined logic for validation, summation, and reporting without invoking a shell or processing arbitrary external scripts.\n- [INDIRECT_PROMPT_INJECTION]: The skill presents a surface for indirect prompt injection as it processes user-provided descriptions of employment contracts and rules.\n - Ingestion points: Data is ingested as JSON via the
check_sidework_rules.pyscript from standard input or local files.\n - Boundary markers: The agent is explicitly instructed in
SKILL.mdto distinguish between confirmed document text and hearsay, and to use specific labels for unverified items.\n - Capability inventory: The skill lacks network access, file-writing permissions, and the ability to execute system commands or external scripts.\n
- Sanitization: All inputs are subjected to strict type validation and schema checks using the
_common.pyutility script.
Audit Metadata