The Agent Skills Directory

[REMOTE_CODE_EXECUTION]: The skill provides installation instructions that involve downloading a binary and extracting it directly into a system-wide binary directory: curl -sL [URL] | tar xz -C /usr/local/bin. This pattern lacks integrity verification, such as checksum or signature checking, which could lead to the execution of malicious code if the source repository or the connection is compromised.
[EXTERNAL_DOWNLOADS]: The skill downloads binaries and source code from github.com/JulienTant/blogwatcher-cli. While this belongs to the skill's author, it is a personal repository not identified as a pre-verified trusted organization or well-known service.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process data from external, untrusted RSS/Atom feeds and websites via HTML scraping. This data is then presented to the agent, creating a risk of indirect prompt injection where malicious instructions could be embedded in feed content.
Ingestion points: Data enters the agent's context through the output of blogwatcher-cli scan and site scraping results.
Boundary markers: No specific delimiters or boundary instructions are used to separate untrusted feed content from other agent instructions.
Capability inventory: The skill uses subprocess execution via blogwatcher-cli and can access URLs using web_extract.
Sanitization: There is no evidence of content sanitization or validation of the feed data before it is processed by the AI.
[COMMAND_EXECUTION]: The skill operates by executing the blogwatcher-cli tool via shell commands for monitoring and managing article states.

blogwatcher