The Agent Skills Directory

[DYNAMIC_EXECUTION]: The skill executes inline Python code via python3 -c to handle file format parsing, specifically for extracting text from .docx files and validating the integrity of .pst archives.
[COMMAND_EXECUTION]: The skill relies on external command-line utilities such as antiword, catdoc, and readpst to process legacy Microsoft Word documents and Outlook data files.
[INDIRECT_PROMPT_INJECTION]:
Ingestion points: The skill processes large volumes of untrusted data from multiple sources, including local filesystem mounts, Dropbox, Backblaze B2, and Gmail archives.
Boundary markers: There are no explicit instructions or delimiters defined to prevent the agent from interpreting instructions that might be embedded within the files being scanned.
Capability inventory: The skill has the ability to read arbitrary files within the allow-listed paths, execute parsing commands, and write the extracted content to the agent's long-term storage (brain pages).
Sanitization: While the skill filters for 'high-value' content and strips HTML tags for display, it does not implement specific sanitization to neutralize potential prompt injection attacks contained in the processed text.

archive-crawler