deeflect

Search Dory memory before factual claims about people, projects, decisions, dates, prior sessions, or cleanup markers; follow search results with exact gets.

Use Dory wake once at the start of a new agent session, after context compaction, or at a real task switch to load bounded hot context with the right profile before substantive work.

Safely write durable Dory memory through semantic or exact-path write surfaces using dry-run, inbox, canonical-write, and purge guards.

Use Dory dreaming after a session closes to distill logs, generate reviewable memory proposals, and apply or reject them safely.

Inspect Dory corpus quality, run maintenance checks, and generate reviewable cleanup or wiki-health reports without mutating memory automatically.

The ritual of atonement. Invoke when the agent (Claude or Codex) commits an unforgivable mistake — deleting the wrong thing, shipping a lie, ignoring an explicit instruction, breaking what worked. Performs a formal apology, draws the agent committing seppuku in ASCII, and records the sin to the global ~/.seppuku ledger so it is never repeated. Also runs when the user types /seppuku.

Use this skill when creating, redesigning, critiquing, reducing, polishing, hardening, or making design decisions for a user-facing interface, brand surface, product UI, component, dashboard, flow, app shell, onboarding, empty state, or visual system. Applies Mies van der Rohe-inspired restraint, the Interface Craft workflow, and anti-generic design discipline. Not for backend-only work.