docs: interface routing, local shell path rewrite, claude CLI, unfilled $ARGUMENTS; release 0.16.1

CHANGELOG.md

@@ -2,6 +2,26 @@
 
 All notable changes to this project will be documented in this file.
 
+## [0.16.1] - 2026-10-03
+
+### Fixed
+- **Where a Cowork task runs, by interface.** 0.16.0 said choosing Cowork in the Chat/Cowork picker
+  starts a cloud session. It runs locally when "Only on this computer" is on and in the cloud when
+  it's off; the newer interface ran in the cloud even with the setting on. `environments.md` now says
+  so and quotes the app's notice on local tasks that new ones can't be started from October 6.
+- **Plugin and skill paths in a local session's shell.** The guide said the VM shell can't resolve
+  the substituted host path. Current Desktop builds rewrite those paths inside shell commands, so the
+  token works there; the guide now warns instead not to hand a path the shell printed to the file
+  tools, which refuse VM paths.
+- **The `claude` CLI is not only in Claude Code.** It was found in a local session's shell, and its
+  absence from the chat runtime was never measured. The README, `SKILL.md` and the
+  `claude-cli-dependency` lint message now say it may be off the PATH outside Claude Code and to gate
+  a `claude -p` step on `command -v claude`.
+
+### Added
+- **`$ARGUMENTS` can arrive unfilled** in a cloud session when the skill is invoked before the
+  session exists. The setup guidance now says to take the input from the user's message then.
+
 ## [0.16.0] - 2026-10-02
 
 ### Fixed

README.md

@@ -30,7 +30,7 @@ The skill that builds skills. Draft one and ship it in a single pass, or run eva
 Anthropic ships a `skill-creator` plugin. It's good, but several parts are broken or missing:
 
 - **Best practices guide included** — 620+ lines of patterns, structural templates, troubleshooting guide, and checklists extracted from Anthropic's [Complete Guide to Building Skills for Claude](https://resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf) and Thariq's [Lessons from Building Claude Code Skills](https://x.com/trq212/status/2024574133011673516). Includes a "Script vs. Instruct" decision framework for when to bundle pre-made scripts vs. keep logic as instructions — covering context window efficiency, reliability, and auditability. The built-in doesn't ship any of this.
-- **A skill that works here can break there, silently** — the `claude` CLI exists only in Claude Code, so a `claude -p` step is dead anywhere else; the Claude app's chat runtime has no sub-agent tool, so parallel eval runs have to collapse to serial there (a Claude app conversation running in a cloud session does have one — route by the tool list, not the product); an agent in a cloud or local session can't serve a local HTTP server and open it, so the eval viewer needs a static build; a third-party import there costs an install on every run, and egress is org-configurable, so a locked-down org can refuse it; and file delivery differs per surface — in a cloud session, writing a file is not delivering it. Structure validation can't see any of that, so there are two checks: `quick_validate` for structure, and a 14-rule portability linter for runtime assumptions — including the two ways compaction loses a skill. Both are stdlib-only, because they have to run inside the sandboxes they lint. Rule ids, flags and exit codes are in [`docs/DEVELOPMENT.md`](docs/DEVELOPMENT.md).
+- **A skill that works here can break there, silently** — the `claude` CLI is on PATH in Claude Code and was found in a local session's shell, but other runtimes are unchecked, so gate a `claude -p` step on `command -v claude`; the Claude app's chat runtime has no sub-agent tool, so parallel eval runs have to collapse to serial there (a Claude app conversation running in a cloud session does have one — route by the tool list, not the product); an agent in a cloud or local session can't serve a local HTTP server and open it, so the eval viewer needs a static build; a third-party import there costs an install on every run, and egress is org-configurable, so a locked-down org can refuse it; and file delivery differs per surface — in a cloud session, writing a file is not delivering it. Structure validation can't see any of that, so there are two checks: `quick_validate` for structure, and a 14-rule portability linter for runtime assumptions — including the two ways compaction loses a skill. Both are stdlib-only, because they have to run inside the sandboxes they lint. Rule ids, flags and exit codes are in [`docs/DEVELOPMENT.md`](docs/DEVELOPMENT.md).
 - **Claude Code runtime docs** — the mechanics most skill authors hit the hard way, read out of the shipping binary rather than inherited from a blog post: overflow of the shared listing budget drops descriptions **per skill, least-recently-used first**, packing first-fit, so full and name-only entries coexist and a long description can lose to a shorter one; `allowed-tools` **grants** permission rather than requesting it; and compaction is a CHARACTER gate, not the documented token one, losing content two different ways — truncation keeps the first 19,900 characters and leaves a marker, while the combined cross-skill cap **zeroes** a skill outright with no marker and no entry. Verified against Claude Code 2.1.222–2.1.251; the listing budget re-verified at 2.1.280. Where the public docs and the binary disagree, [the reference](skill-creator-plus/skills/skill-creator-plus/references/official-guide-patterns.md) says so and shows which one shipped.
 
 The eval viewer, description optimizer and benchmarking fixes are in the table above; see the [CHANGELOG](CHANGELOG.md) for the full list.

VERSION

@@ -1 +1 @@
-0.16.0
+0.16.1