run-deep-research
Pass
Audited by Gen Agent Trust Hub on Oct 9, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill's primary function is to ingest and process untrusted data from the web via subagents. This creates a potential surface for indirect prompt injection. However, the skill incorporates significant mitigations, including mandatory 'verbatim quotations' from primary sources, source ledgers (
09-sources.md), and a 6-dimension evaluation audit that requires evidentiary rigor. - [COMMAND_EXECUTION]: The skill provides instructions for subagents to use
curlas a fallback mechanism for web fetching if primary platform tools fail. This is a legitimate utility for the research task and is not used for malicious purposes such as pipe-to-bash execution or credential exfiltration. - [EXTERNAL_DOWNLOADS]: The documentation includes standard installation commands using
npxandprsthat reference the author's public GitHub repository (yigitkonur/skills-by-yigitkonur). These are standard for AI agent skill distribution and target the vendor's own verified resources. - [SAFE]: The skill includes scaffolding scripts (
scripts/init-corpus.sh) and link-checking utility scripts (Ruby snippets in documentation) which are benign and intended for project maintenance and integrity verification.
Audit Metadata