disable-model-invocation for dump-session and warm-handoff;
user-invocable: false for describing-changes, galahad,
self-improving-skills, and working-in-the-open.

Assisted-by: Claude Opus 4.6 via pi

diff --git a/ai/claude/skills/describing-changes/SKILL.md b/ai/claude/skills/describing-changes/SKILL.md
index 441c54fe..d5cb8827 100644
--- a/ai/claude/skills/describing-changes/SKILL.md
+++ b/ai/claude/skills/describing-changes/SKILL.md
@@ -1,6 +1,7 @@
 ---
 name: describing-changes
 description: Use when drafting any explanation of changes (commit messages, PRs, code reviews, docs)
+user-invocable: false
 ---
 
 # Describing Changes
diff --git a/ai/claude/skills/dump-session/SKILL.md b/ai/claude/skills/dump-session/SKILL.md
index 3a728f21..f31a5f36 100644
--- a/ai/claude/skills/dump-session/SKILL.md
+++ b/ai/claude/skills/dump-session/SKILL.md
@@ -1,6 +1,7 @@
 ---
 name: dump-session
 description: This skill should be used when the user asks to "dump session", "export session", "save session", "/dump-session", "export conversation", "save this conversation", or "write session to markdown". Exports the current Claude Code session as Markdown and HTML files with a summary.
+disable-model-invocation: true
 ---
 
 # Dump Session
diff --git a/ai/claude/skills/galahad/SKILL.md b/ai/claude/skills/galahad/SKILL.md
index 29d9ad56..dbd136ec 100644
--- a/ai/claude/skills/galahad/SKILL.md
+++ b/ai/claude/skills/galahad/SKILL.md
@@ -2,6 +2,7 @@
 name: galahad
 description: Use when running tests, fixing type errors, or improving coverage - prioritizes 100% green checks as a trustworthy signal
 source: https://github.com/lambdamechanic/skills/blob/main/galahad/SKILL.md
+user-invocable: false
 ---
 
 # Coding Agent Quality Rules (Galahad Principle)
diff --git a/ai/claude/skills/self-improving-skills/SKILL.md b/ai/claude/skills/self-improving-skills/SKILL.md
index b76d83d3..57622ba0 100644
--- a/ai/claude/skills/self-improving-skills/SKILL.md
+++ b/ai/claude/skills/self-improving-skills/SKILL.md
@@ -1,6 +1,7 @@
 ---
 name: self-improving-skills
 description: Use when a loaded skill is incomplete or sparse, or when you discover patterns, gotchas, or commands worth capturing in an existing skill during normal work.
+user-invocable: false
 ---
 
 # Self-Improving Skills
diff --git a/ai/claude/skills/warm-handoff/SKILL.md b/ai/claude/skills/warm-handoff/SKILL.md
index 572d2190..777e416a 100644
--- a/ai/claude/skills/warm-handoff/SKILL.md
+++ b/ai/claude/skills/warm-handoff/SKILL.md
@@ -2,6 +2,7 @@
 name: warm-handoff
 description: Use when the user asks to "warm handoff", "/warm-handoff", "hand off this thread", or "redirect this to". Reads a Slack thread and proposes a warm handoff message for the target channel.
 argument-hint: <slack-thread-url> <#target-channel>
+disable-model-invocation: true
 ---
 
 # Warm Handoff
diff --git a/ai/claude/skills/working-in-the-open/SKILL.md b/ai/claude/skills/working-in-the-open/SKILL.md
index 4c467711..a93486ce 100644
--- a/ai/claude/skills/working-in-the-open/SKILL.md
+++ b/ai/claude/skills/working-in-the-open/SKILL.md
@@ -1,6 +1,7 @@
 ---
 name: working-in-the-open
 description: Use when doing work on a tracked issue (GitHub, Jira, Linear, etc) — documents decisions, blockers, and summaries as comments on the issue for historical record
+user-invocable: false
 ---
 
 # Working in the Open