The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to execute multiple shell commands, including curl, sed, grep, and tr. It interpolates user-provided URLs directly into these commands, which could lead to command injection if the execution environment does not properly sanitize shell arguments.
[EXTERNAL_DOWNLOADS]: The skill recommends installing third-party packages at runtime using npm install -g @mozilla/readability-cli, npm install -g reader-cli, and pip3 install trafilatura. These are external dependencies that modify the agent's execution environment.
[INDIRECT_PROMPT_INJECTION]: The skill possesses a surface for indirect prompt injection by design.
Ingestion points: Content is fetched from arbitrary external URLs provided by the user using curl, reader, or trafilatura (SKILL.md).
Boundary markers: The skill does not use specific delimiters or instructions to the agent to ignore potentially malicious commands embedded within the extracted article text.
Capability inventory: The agent has the ability to execute shell commands, install software, write to the filesystem, and preview file contents.
Sanitization: While filenames are cleaned for filesystem compatibility, the article content itself is not sanitized or filtered for instructions that might target the agent.
[DYNAMIC_EXECUTION]: The skill uses python3 -c to execute an inline Python script for HTML parsing. While the script is static within the skill instructions, it dynamically processes untrusted data fetched from the network.

content-harvest