From daa3fcd14b16581a36d9e0f57195a09e46a8c7c2 Mon Sep 17 00:00:00 2001 From: Claude Date: Wed, 15 Apr 2026 17:32:13 +0000 Subject: [PATCH] Add psake skill note to 5.0.0 release post Cross-link the psake Agent Skill (via the LLM tools marketplace) from the 5.0.0 announcement so readers using Claude or Copilot know the skill covers the new declarative syntax, caching, and structured output features. https://claude.ai/code/session_01EnA7YLp8eQiRpBVXridfcx --- blog/2026-04-15-psake-5-0-0-released.md | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/blog/2026-04-15-psake-5-0-0-released.md b/blog/2026-04-15-psake-5-0-0-released.md index f823822..101ed27 100644 --- a/blog/2026-04-15-psake-5-0-0-released.md +++ b/blog/2026-04-15-psake-5-0-0-released.md @@ -111,6 +111,12 @@ if (-not $result.Success) { Rather than parsing wall-of-text build output, the agent can check `$result.Success`, iterate `$result.Tasks`, and read `$_.Error` directly—structured data that maps cleanly to a tool call response. Pair this with `-OutputFormat JSON` if you need to pass the result across a process boundary or into a prompt. +:::tip Using psake with an AI coding assistant? + +The [psake Agent Skill](/blog/introducing-psake-agent-skill) has been updated for 5.0 and is available through the [LLM tools marketplace](https://github.com/psake/psake-llm-tools). Install it into Claude, GitHub Copilot, or any Agent Skills–compatible client to give your assistant first-class knowledge of the new declarative syntax, caching, structured output, and migration patterns covered in this post. + +::: + ## New Testing APIs Two new functions let you inspect and test your build scripts without running a full build: