The Agent Skills Directory

[COMMAND_EXECUTION]: The SKILL.md file contains instructions to execute the research script using the pattern python3 ./scripts/last30days.py "$ARGUMENTS". Passing user-supplied input directly into a shell command via $ARGUMENTS creates a risk of command injection if the underlying platform does not sanitize the input, potentially allowing an attacker to execute arbitrary shell commands.
[INDIRECT_PROMPT_INJECTION]:
Ingestion points: The skill retrieves data from external, untrusted sources including Reddit threads (scripts/lib/reddit_enrich.py), X (Twitter) posts (scripts/lib/xai_x.py), and general web pages (scripts/lib/websearch.py).
Boundary markers: The instructions do not define clear delimiters or "ignore embedded instructions" warnings for the agent when processing this external content.
Capability inventory: The skill possesses capabilities for outbound network communication (urllib.request) and local file system writes (scripts/lib/render.py, scripts/lib/cache.py).
Sanitization: There is no evidence of sanitization or escaping of the ingested content before it is synthesized into a "Prompt Pack" for the user, meaning malicious instructions embedded in social media posts could influence the agent's output.
[EXTERNAL_DOWNLOADS]: The skill makes network requests to api.openai.com, api.x.ai, and reddit.com. These requests are necessary for the skill's stated purpose of aggregating social media discussions and use official endpoints for well-known services.

last30days