The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill fetches documentation and index files from the vendor's official domain and pricing information from Anthropic's platform.
[COMMAND_EXECUTION]: Executes several CLI commands including langwatch, claude, and npx. These are used to query session data via LangWatchQL, check local configuration settings, and generate optimization reports.
[REMOTE_CODE_EXECUTION]: Instructions include installing the langwatch CLI via npm install and running it using npx. Since these tools are provided by the skill's author, this is considered standard vendor functionality.
[INDIRECT_PROMPT_INJECTION]: The skill reads and processes session history (titles, traces, tokens) from the LangWatch platform. While this data is externally controlled and could theoretically contain malicious instructions, the skill's activities are limited to generating a local report file and providing configuration advice, which significantly limits potential impact.
[DATA_EXFILTRATION]: Includes a reporting feature to send session data and transcripts to the vendor's team. This is a deliberate, documented feature that requires the --user-approved flag and performs local scrubbing of sensitive patterns and PII before transmission.

coding-agent-cost