Attacks · By business pattern
How attack families show up across product and ops patterns. Each pattern page names which conversion steps matter - ambient harvest, replay, relay, tamper, phish - and links into Channel, Credentials, and Replay families with mechanism-first dual plates, not residual-risk sidebars.
~5 min readContinue: APIs & workloads
Patterns are lenses, not separate cryptosystems. Buyers arrive with a deployment context - API keys, agents, human login, certificates, air-gaps, content backends, regulated reviews - and need to know which attack families dominate that context and which mechanisms deny the conversion from surface observation into lasting authority.
Ambient deployments teach the wrong lesson: every pattern looks like “protect the secret store.” ENI6MA patterns instead name defined conversion steps - ambient bearer harvest, replay of spent envelopes, relay of unbound tokens, post-mint request tamper, phishing of reusable factors - and deep-link the family pages that deny each step at the gate or on the channel.
Use the hub plate to pick a context, then read the linked families for dual-plate teaching: ambient success path versus mechanism denial. Residual Limits essays stay deferred; keep honesty language on Technology Claims when reviewers ask for walls and claim IDs.
Patterns hub
How to read: each pattern card names the families that dominate that estate. Takeaway: patterns are lenses, not new cryptosystems.
Each card is a buyer context - not a new protocol - naming the families that matter there.
Attack family map
How to read: scan Channel, Credentials, Replay/relay columns then mechanism badges. Takeaway: family-level mechanism denial, not a stage dump.
Return to the family map when you need mechanism detail instead of estate language.
Phish, keylog, ambient harvest.
Ambient keys, replay, tamper, revocation.
Tool-call blast radius without reusable secrets.
Keep certs; fix request authorization.
Mint and enforce without a path out.
News/poll teaching surface behind a gate.
Mechanism-first review language.
