The Agent Skills Directory

[COMMAND_EXECUTION]: The script prompt-engineer-toolkit/scripts/prompt_tester.py utilizes subprocess.run() to execute shell commands provided by the user through the --runner-cmd argument. This allows for arbitrary command execution on the host system.
[DYNAMIC_EXECUTION]: The script prompt-engineer-toolkit/scripts/prompt_tester.py dynamically constructs commands using user-provided templates. Furthermore, x-twitter-growth/scripts/growth_tracker.py constructs file paths using unsanitized user input (the Twitter handle), which could lead to path traversal vulnerabilities when writing data snapshots.
[EXTERNAL_DOWNLOADS]: The site-architecture/scripts/sitemap_analyzer.py utility uses urllib.request.urlopen() to fetch sitemap files from user-supplied external URLs, potentially allowing for server-side request forgery (SSRF) or downloading malicious content.
[INDIRECT_PROMPT_INJECTION]: Multiple analysis scripts, including copy-editing/scripts/readability_scorer.py and content-humanizer/scripts/humanizer_scorer.py, ingest and analyze untrusted text documents. The absence of strict boundary markers or input sanitization, combined with the agent's capability to execute commands via the toolkit, creates a surface for indirect prompt injection attacks where malicious content in the analyzed documents could attempt to hijack the agent's execution flow.

marketing-skills