<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>MCP Servers on processkit</title><link>https://projectious-work.github.io/processkit/v1.x/docs/mcp-servers/</link><description>Recent content in MCP Servers on processkit</description><generator>Hugo</generator><language>en</language><atom:link href="https://projectious-work.github.io/processkit/v1.x/docs/mcp-servers/index.xml" rel="self" type="application/rss+xml"/><item><title>Harness Compatibility</title><link>https://projectious-work.github.io/processkit/v1.x/docs/mcp-servers/harness-compatibility/</link><pubDate>Mon, 01 Jan 0001 00:00:00 +0000</pubDate><guid>https://projectious-work.github.io/processkit/v1.x/docs/mcp-servers/harness-compatibility/</guid><description>&lt;p&gt;processkit&amp;rsquo;s MCP servers are provider-neutral Python programs. They do
not require aibox at runtime. aibox can install processkit, merge MCP
configuration, pre-authorize processkit tools where a harness supports
that, and supervise a managed devcontainer. Those are convenience and
lifecycle features; they are not a processkit dependency.&lt;/p&gt;
&lt;p&gt;The alpha.5 installer can project managed Codex and Claude configuration.
It owns only declared processkit keys and preserves unrelated harness settings
during install, update, and uninstall. Restart the harness after installation
so it reloads the projection.&lt;/p&gt;</description></item><item><title>Claude Code</title><link>https://projectious-work.github.io/processkit/v1.x/docs/mcp-servers/claude-code/</link><pubDate>Mon, 01 Jan 0001 00:00:00 +0000</pubDate><guid>https://projectious-work.github.io/processkit/v1.x/docs/mcp-servers/claude-code/</guid><description>&lt;blockquote&gt;
&lt;p&gt;&lt;strong&gt;v1 alpha note:&lt;/strong&gt; Install with &lt;code&gt;--harness claude&lt;/code&gt; to create the managed
projection. The installer preserves unrelated Claude configuration. The
direct &lt;code&gt;uv&lt;/code&gt; gateway remains the development and compatibility fallback.&lt;/p&gt;
&lt;/blockquote&gt;
&lt;p&gt;This document captures how processkit surfaces itself inside the Claude
Code harness: which payloads land at session start vs. each turn, which
hooks fire, and which Claude Code settings we recommend for any
processkit project.&lt;/p&gt;
&lt;p&gt;WorkItem: BACK-20260509_1317-DaringRaven (issue #19). Companion file:
&lt;a href="settings.example.json"&gt;&lt;code&gt;settings.example.json&lt;/code&gt;&lt;/a&gt;
.&lt;/p&gt;
&lt;h2 id="what-the-per-turn-hook-injects"&gt;What the per-turn hook injects&lt;/h2&gt;
&lt;p&gt;&lt;code&gt;scripts/emit_compliance_contract.py&lt;/code&gt; runs as &lt;strong&gt;both&lt;/strong&gt; a &lt;code&gt;SessionStart&lt;/code&gt;
and a &lt;code&gt;UserPromptSubmit&lt;/code&gt; hook (wired in
&lt;code&gt;/workspace/.claude/settings.json&lt;/code&gt; under &lt;code&gt;hooks.SessionStart&lt;/code&gt; and
&lt;code&gt;hooks.UserPromptSubmit&lt;/code&gt;). It now emits two different payloads:&lt;/p&gt;</description></item></channel></rss>