social-engineeringlisted
Install: claude install-skill noahfranklin/gideon
# Social Engineering (Authorized Assessment)
Methodology for **sanctioned** human-layer testing that measures and improves an organization's resilience to social attacks. This is about strengthening the workforce — every action requires organizational authorization and follows agreed rules that protect employees.
## 0. Authorization & ethics gate (mandatory)
- [ ] Written authorization from an executive sponsor who can consent on the org's behalf.
- [ ] Defined scope: which groups, channels (email/phone/SMS), and pretext limits.
- [ ] **Employee-protection rules:** no capturing real passwords into attacker-held stores, no lasting harm, no coercion/harassment, no sensitive-topic lures (medical, layoffs, etc.) unless explicitly agreed.
- [ ] Data handling: PII minimization, secure storage, defined retention/deletion.
- [ ] Legal/HR review and a debrief/awareness plan for participants.
This skill is for programs where the target *organization* consents. It must not be used to deceive individuals outside such a program.
## 1. Recon (for realism, within scope)
- Public info that a real attacker would use: email formats, org structure, tech/vendors, current events (from `recon-osint`). Keep pretexts plausible but non-harmful.
## 2. Campaign design
- **Objective:** measure click / credential-submit / report rates, or test a specific control (e.g., MFA prompt handling), not to "catch" individuals.
- **Pretext:** believable, within agreed sensitivity limits.
- **Landing pages:**