From 72563aede0d7ddaab38774894f13262fcca0b884 Mon Sep 17 00:00:00 2001
From: Amolith <amolith@secluded.site>
Date: Fri, 24 Apr 2026 17:03:02 -0600
Subject: [PATCH] readme: update

---
 README.md                                     | 29 ++++++++++++++-----
 .../updating-llm-client-model-lists/SKILL.md  |  7 ++++-
 2 files changed, 27 insertions(+), 9 deletions(-)

diff --git a/README.md b/README.md
index 23a2b2309a10e22cbd288c0f00dab7bc0fa6ef03..7a1a50738eebcc6afe6a49bf82bbe721cdfd5256 100644
--- a/README.md
+++ b/README.md
@@ -109,6 +109,9 @@ token count, plus overall metadata usage. I've used and tested them most with
 - [using-jujutsu](skills/using-jujutsu/SKILL.md): Guides [jujutsu] (jj)
   workflows, covering commit selection, change creation, diff reading, and
   common pitfalls that trip up agents.
+- [using-silverbullet](skills/using-silverbullet/SKILL.md): Manages notes in
+  [SilverBullet] via the CLI and Space Lua. Covers searching, reading, writing,
+  and editing pages, plus journal and task workflows.
 - [working-with-tmux](skills/working-with-tmux/SKILL.md): Spawns and manages
   background processes via tmux. Creates windows, sends commands, and captures
   output without blocking the main session. Useful for servers and long tasks.
@@ -136,6 +139,7 @@ token count, plus overall metadata usage. I've used and tested them most with
 [gum]: https://github.com/charmbracelet/gum
 [exe.dev]: https://exe.dev
 [jujutsu]: https://jj-vcs.github.io/jj/latest/
+[SilverBullet]: https://silverbullet.md
 
 ## Installation
 
@@ -508,9 +512,9 @@ Token breakdown:
 Token breakdown:
   Name:           13 tokens
   Description:    51 tokens
-  Body:          997 tokens (126 lines)
+  Body:         1124 tokens (131 lines)
   ───────────────────────────────────────────────
-  Total:        1061 tokens
+  Total:        1188 tokens
 
 === using-exe-dev ===
 
@@ -530,14 +534,23 @@ Token breakdown:
   ───────────────────────────────────────────────
   Total:         799 tokens
 
+=== using-silverbullet ===
+
+Token breakdown:
+  Name:            8 tokens
+  Description:    45 tokens
+  Body:         1001 tokens (118 lines)
+  ───────────────────────────────────────────────
+  Total:        1054 tokens
+
 === working-with-tmux ===
 
 Token breakdown:
   Name:            8 tokens
   Description:    32 tokens
-  Body:          909 tokens (128 lines)
+  Body:          916 tokens (129 lines)
   ───────────────────────────────────────────────
-  Total:         949 tokens
+  Total:         956 tokens
 
 === writing-git-tags ===
 
@@ -578,10 +591,10 @@ Token breakdown:
 SUMMARY
 ============================================================
 
-Skills: 31
-Metadata: 1949 tokens
-Combined bodies: 32856 tokens
-Overall: 88318 tokens
+Skills: 32
+Metadata: 2002 tokens
+Combined bodies: 33991 tokens
+Overall: 89506 tokens
 Validation errors: 0
 
 Largest skills (by total tokens):
diff --git a/skills/updating-llm-client-model-lists/SKILL.md b/skills/updating-llm-client-model-lists/SKILL.md
index 93f6f889544bf7307382aa4c041d367b9397abd9..a9d924495a87ea616287420272b6fe4bfd2f3568 100644
--- a/skills/updating-llm-client-model-lists/SKILL.md
+++ b/skills/updating-llm-client-model-lists/SKILL.md
@@ -5,6 +5,10 @@ description: Synchronizes model configurations across Zed, Crush, Octofriend, an
 
 Updates four LLM client configs from the Plexus proxy's available models.
 
+## max_completion_tokens
+
+Always use `floor(context_length * 0.2)` as the max_completion_tokens value for all clients. Do NOT use the `top_provider.max_completion_tokens` field from the API — some models report their entire context window as max completion tokens, which is not a sensible default. 20% of context length is a more practical cap.
+
 ## Network reachability
 
 1. Ping `harp-willow.snowy-egret.ts.net` once
@@ -40,6 +44,7 @@ This should provide all the info you need in a table. The data in this table is
 
 - Always exclude `nomic-embed-text-v1.5` (embedding model, not supported by any client)
 - Exclude `diff-apply` and `fix-json` from Zed, Crush, and Pi because it's only supported by Octofriend
+- Exclude models with no `context_length` field (e.g. voxtral-small, mistral-ocr) unless the user provides values manually
 - Omit special characters from display names. For example, "MiniMax-M2.7" should become "MiniMax M2.7". "nemotron-3-super" becomes "Nemotron 3 Super".
 
 ## Update Zed config
@@ -128,4 +133,4 @@ Update `providers.plexus.models[]`. Each entry:
 ```
 
 After editing: `chezmoi apply ~/.config/pi/models.json`
-STOP if chezmoi has any output other than success.
+NEVER use `--force` with chezmoi apply. If it reports the file has changed since last write, STOP and inform the user — they may have local changes not yet added to chezmoi.