The Agent Skills Directory

[PROMPT_INJECTION]: The skill ingests data from external websites via WebFetch and mcp__chrome-devtools__. This creates a surface for indirect prompt injection where instructions hidden in the scraped content could influence the agent's output. 1. Ingestion points: External URLs defined in sources.json. 2. Boundary markers: The sub-agents are instructed to return only JSON, but there are no markers in the main orchestrator prompt. 3. Capability inventory: Write access to the filesystem and restricted Bash commands. 4. Sanitization: No explicit content sanitization for instructions is implemented.
[COMMAND_EXECUTION]: The skill requests Bash access but restricts it to safe whitelisted commands (mkdir, date, ls), which significantly reduces the risk of command injection.

daily-news-report