read-deleted-pages
Pass
Audited by Gen Agent Trust Hub on Aug 24, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill provides numerous examples using standard shell utilities such as
curl,wget,jq,grep, andawkto interact with web archive APIs and process metadata. - [EXTERNAL_DOWNLOADS]: The skill facilitates data retrieval from well-known public archives including the Internet Archive (
web.archive.org),archive.today, andcommoncrawl.org. These sources are recognized as established services appropriate for the skill's OSINT purpose. - [PROMPT_INJECTION]: The skill's instructions to fetch and analyze content from external websites create an attack surface for indirect prompt injection (Category 8).
- Ingestion points: External data is ingested through
curlandwgetcommands targeting third-party archives and target domains (found inSKILL.mdandreference/cdx-cookbook.md). - Boundary markers: The skill does not provide explicit delimiters or instructions to the agent to disregard potential instructions within the retrieved content.
- Capability inventory: The agent is guided to perform network operations, write data to files (e.g.,
wget --warc-file), and process results using various command-line tools. - Sanitization: No content sanitization or validation mechanisms are included to handle untrusted external data before it is processed by the agent.
Audit Metadata