The Agent Skills Directory

[COMMAND_EXECUTION]: The skill includes a Python script scripts/plan_route.py. Analysis shows it is a static utility for calculating distances and travel times between pre-defined Israeli cities using an internal dictionary. It does not use any dangerous functions like os.system, subprocess, or eval, and has no external dependencies.
[EXTERNAL_DOWNLOADS]: The skill mentions and provides links to several Recommended MCP Servers (e.g., israel-railways, openbus) located at agentskills.co.il. These are presented as optional external tools for live data. The links are consistent with the skill's stated purpose and author.
[DATA_EXFILTRATION]: No evidence of unauthorized data collection or exfiltration was found. The skill does not perform network operations or access sensitive local files. The URLs provided in the documentation and evidence.json point to official Israeli government domains (pti.org.il, parks.org.il), transport providers (egged.co.il, ravkavonline.co.il), and reputable news or information sites (wikipedia.org, moovitapp.com).
[OBFUSCATION]: No malicious obfuscation techniques such as Base64-encoded payloads, zero-width characters, or homoglyph attacks were detected. Standard URL percent-encoding for Hebrew characters is used in reference links, which is expected behavior.
[PROMPT_INJECTION]: The instructions in SKILL.md and SKILL_HE.md are purely informational and do not attempt to override the AI agent's safety protocols or core behavior. The examples provided are standard use cases for the skill's functionality.
[INDIRECT_PROMPT_INJECTION]: While the skill processes user travel queries, it lacks high-privilege capabilities (such as writing to the filesystem or performing arbitrary network requests) that could be exploited via indirect injection. The risk is assessed as safe within the intended operating environment.

israeli-travel-planner