The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses the Bash tool to execute its own local Python orchestration script (last30days.py) to manage the research pipeline, including parallel API requests and data normalization.
[EXTERNAL_DOWNLOADS]: The skill performs network operations to fetch research data from well-known services. It communicates with api.openai.com (for Reddit discovery via the Responses API), api.x.ai (for X search), and www.reddit.com (for thread enrichment). These downloads are necessary for the skill's primary function and target recognized service providers.
[INDIRECT_PROMPT_INJECTION]: The skill has a vulnerability surface for indirect prompt injection because it ingests untrusted text from social media and instructs the agent to synthesize this data into copy-paste-ready prompts for the user. Malicious content within social media posts could potentially influence the agent's output.
Ingestion points: Untrusted social media text is retrieved through API responses in scripts/lib/openai_reddit.py, scripts/lib/xai_x.py, and scripts/lib/reddit_enrich.py.
Boundary markers: The instructions in SKILL.md lack explicit delimiters or warnings to ignore instructions that might be embedded within the ingested social media content.
Capability inventory: The skill possesses significant capabilities, including Bash, Read, and Write tool permissions, which could be relevant if a successful injection were to occur.
Sanitization: The skill does not implement specific sanitization or filtering to detect or strip instructions from the social media data before it is processed by the model for synthesis.

last30days