The Agent Skills Directory

[REMOTE_CODE_EXECUTION]: The skill documentation includes an installation command (curl -fsSL ... | bash) that pipes a script from a remote GitHub repository (xdevplatform/xurl) directly into the system shell. This pattern allows for unverified code execution on the host machine.
[EXTERNAL_DOWNLOADS]: The skill requires downloading software from external package managers and repositories, including Homebrew (xdevplatform/tap/xurl), NPM (@xdevplatform/xurl), and Go (github.com/xdevplatform/xurl). These dependencies are managed by a third-party organization (xdevplatform) not recognized as a trusted vendor.
[DATA_EXFILTRATION]: The tool manages sensitive X API tokens stored locally in ~/.xurl. While the skill includes instructions to prevent the agent from printing these tokens, the underlying tool has broad network access capabilities and raw API execution features that could be abused to transmit sensitive information if the agent is compromised.
[PROMPT_INJECTION]: The skill operates as a conduit for untrusted data from the X API, such as search results, mentions, and direct messages. This creates a significant surface for indirect prompt injection, where malicious instructions embedded in social media content could influence the agent's logic.
Ingestion points: API output from timeline, mentions, search, dms, and read commands in SKILL.md.
Boundary markers: The skill lacks explicit instructions for the agent to treat API output as untrusted data or use delimiters to prevent command following.
Capability inventory: The skill possesses write capabilities (xurl post, xurl dm), media upload functions, and raw API access (xurl -X POST) as documented in SKILL.md.
Sanitization: No evidence of sanitization, filtering, or validation of external content before processing in the instructions.

xurl