x-bookmark-quote-posts
Pass
Audited by Gen Agent Trust Hub on Aug 30, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests data from external X/Twitter bookmarks which constitutes untrusted input. If a bookmarked post contains malicious instructions, the agent might attempt to follow them while generating drafts.
- Ingestion points: The skill collects data from
https://x.com/i/bookmarksand follows source status URLs as defined inSKILL.md. - Boundary markers: There are no explicit delimiters or instructions provided to the agent to separate the source text from the system instructions or to ignore embedded commands within the content.
- Capability inventory: The skill is authorized to write to the file system (e.g.,
data/x-growth/bookmark-quote-posts/YYYY-MM-DD.md) and execute shell commands (git commit). - Sanitization: The instructions do not specify any sanitization, filtering, or validation of the text extracted from the bookmarks before it is processed.
- [COMMAND_EXECUTION]: The skill utilizes shell commands to manage a local repository and verify changes.
- Evidence:
SKILL.mdcontains instructions to executegit status --short,git diff --check, andgit committo manage the content queue.
Audit Metadata