Blog
Security Research··1 min read

Technical Advisory: Hardening the Phase 4 Acquisition Path

How we hardened public acquisition endpoints with bounded abuse controls, deterministic event taxonomy, and evidence-linked workflow handoff.

Scope

This advisory documents the Phase 4 acquisition hardening pass and the controls currently enforced in production.

Control set

1) Public-safe acquisition endpoint

Acquisition scans are exposed through a public-safe surface:

  • POST /api/public/acquisition/scan
  • GET /api/public/acquisition/results/{scan_id}

The response intentionally limits detail to trust score, top finding summaries, and shareable trust-card links.

2) Abuse control

Public acquisition scans are bounded by a sliding-window rate limiter on caller IP.

Key behaviors:

  • Per-IP rate limiting on quick scan submission
  • Explicit 429 with retry metadata when limit is exceeded
  • No bypass path through client-side state

3) Share loop and upgrade handoff

Every acquisition scan emits:

  • Public trust-card URL (/acquire/{scan_id})
  • Workflow deep link (/dashboard/workflow)
  • Upgrade CTA with return path to workflow

This keeps acquisition and investigation flows connected without removing existing production paths.

4) Conversion-event contract

Client instrumentation uses a fixed event taxonomy, including:

  • quick_scan_started
  • quick_scan_completed
  • first_scan
  • trust_card_viewed
  • trust_card_shared
  • badge_copied
  • workflow and upgrade transition events

Critical events enforce identity requirements server-side for reliable conversion math.

Current watch items

  • Conversion event stability remains under active verification, with a production stability patch in progress.
  • Acquisition gate decisions are reviewed weekly to ensure keep, change, or remove actions are evidence-driven.

Why this matters

Security tooling adoption starts at trust. The acquisition path is now constrained, measurable, and linked to investigation workflows rather than isolated scoring pages.

Vigile Team

Security research and engineering at Vigile. Building tools to protect developers from malicious AI agent tools.