akemi-updatelisted

Sync the graph with recent code changes. Run this after any edit session. ## Steps 1. If `.akemi/.index-stale` exists: `bash .akemi/scripts/rebuild-index.sh`. Read `.akemi/graph/index.yaml` for current state. List changed files (`git status --short` or from context). 2. Created files: pick the kind (file, class, interface, function, test, config), build the ID `<prefix>-<kebab-case-name>`, fill `.akemi/templates/node/<kind>.yaml` (id, name, path, language, refs: `part_of` module, `implements`, `extends`, `depends_on`, `uses_technology`), write to `.akemi/graph/nodes/<kind>/<id>.yaml`. 3. Modified files: find the node via the index path lookup; update refs to match real imports/deps and set `updated` to today. No node found: create one as in step 2. 4. Moved/renamed files: update the node's `path`, keep the same ID. 5. Deleted files: set the node's `status: deprecated`. Never delete node files; never reuse the ID. 6. Fix cross-refs in affected nodes (add/remove `depends_on` etc.). Every `refs[].to` must exist in the index. 7. Work item linkage: if the change completes a task, update the task node status; if a user flow changed, update the journey YAML in `.akemi/journeys/`. 8. Rebuild and check: ```bash bash .akemi/scripts/rebuild-index.sh bash .akemi/scripts/rebuild-views.sh bash .akemi/scripts/validate.sh ``` FAIL lines: fix the named node YAML, re-run, max 3 attempts, then report the remaining FAIL output verbatim and stop. ## Rules - Never e