memex-opslisted

# Operate Memex safely Prefer read-only diagnosis. Require explicit scope and confirmation for destructive operations. ## Host paths - Codex: derive the plugin root from this SKILL.md path by removing `/skills/memex-ops/SKILL.md`; never derive it from the current working directory. - Claude Code: use `${CLAUDE_PLUGIN_ROOT}` as the plugin root. - Define `<cli>` as `node "<plugin-root>/dist/cli.js"` for Codex or `node "${CLAUDE_PLUGIN_ROOT}/dist/cli.js"` for Claude Code. ## Preconditions - Resolve mode/root and requested operation before running tools. - Start with `doctor` for recovery, path, state, retention, secret, or schedule concerns. - Require explicit user confirmation immediately before `purge` or marketplace/plugin removal. - Never infer destructive scope from conversation history. ## Procedure 1. Run `doctor --mode <mode> --root <root> --json` and classify each finding as healthy, warning, recoverable error, or blocker. 2. For recovery, apply only the action named by the doctor result, then rerun doctor before any other mutation. 3. For schedules, call `schedule` with one host-neutral intent and verify idempotent stored state; host automation creation remains a separate authorized action. 4. For purge, show exact scope (`raw-data`, `schedules`, or `personal-wiki`), paths, and consequences; obtain confirmation, call `purge` once, then verify targets without following symlinks. 5. For uninstall, follow `README.md`, remove the host plugin before optionally removi