setup-jd-solanki-skills
Pass
Audited by Gen Agent Trust Hub on Oct 13, 2026
Risk Level: SAFEPERSISTENCEEXTERNAL_DOWNLOADSPRIVILEGE_ESCALATIONPROMPT_INJECTION
Full Analysis
- [PERSISTENCE]: The skill modifies global agent configuration files, specifically
~/.codex/AGENTS.mdand~/.claude/CLAUDE.md. These modifications persist across future agent sessions on the host machine, mandating long-term behavioral changes such as the automatic loading of specific skills. - [EXTERNAL_DOWNLOADS]: The skill references a collection of third-party repositories and websites for additional agent capabilities, including official repositories from well-known services such as Cloudflare and Vercel Labs.
- [PRIVILEGE_ESCALATION]: It defines a specific workflow for executing
sudocommands, instructing the agent to use theherdrtool to manage elevated privileges and password entry. - [PROMPT_INJECTION]: The instructions employ highly prescriptive language ("MUST", "BLOCKING gate", "violation") to enforce a mandatory "Triage" protocol that overrides how the agent normally processes and reads reference files.
Audit Metadata