chore(memory): flush runtime memories and promote managed skills
Memory-file flush for the dotfiles-deploy session with the blankpad admin credential scrubbed to a placeholder; sessions transcripts untracked (gitignored) since harness transcripts can carry raw secrets. Managed skills: command-center-dotfiles-import, ssh-fish-host-remote-exec, stash-supersession-audit-before-drop.
This commit is contained in:
1 parent
2db711964b
commit
35b871162c
27 files changed
+180
-67
No files matched your search
@@ -1 +1 @@
|
||||
Frank's OMP harness environment on CachyOS/niri Wayland. Dual model config files (`~/.pi/agent/models.json` & `~/.omp/agent/models.yml`) require lock-step editing; context uses Ki-style multiples. `frank-default-small` (dankserver:7770) and `frank-default-large` (dankclaw:7770) run via llama.cpp; large model is Qwen3.8-Flash-Next STRIX_LEAN using ngram-mod speculative decoding. NPU (gemma4-it) on port 7771 strictly requires PNG/JPEG over WebP. Git workflow on `dank/command-center` requires `git stash` before `checkout main` due to dirty `agent/memories/`. NRI Wayland compositor supports hot-reload via `SIGUSR1` or `niri-msg`. Force reboot resolves nvidia-modeset semaphore deadlocks but interrupts SSH.
|
||||
Current repo state (`~/.omp/main`) is authoritative. When editing models, update both `models.json` and `models.yml` in lock-step using Ki-style context (200k = 204800). Use `git stash` before checkout to avoid memory file conflicts. For NPU/FLM tasks, always convert WebP to PNG/JPEG. OMP agents use `APPEND_SYSTEM.md` and `task` tool. Deploy dotfiles with `deploy.sh` (sync/adopt modes); secrets are environment-managed. Niri compositor reloads without full session restart. NPU inference on dankclaw port 7770/7771.
|
||||
Reference in new issue
Block a user