Claude Code mod
self-compact
Compaction at task boundaries: the model can ask to compact once its turn ends. Above a fill threshold it is asked whether to, after a turn decision-model judges a finished task, or, without decision-model, with each prompt.
Install
/plugin marketplace add kzarzycki/claude-mods
/plugin install self-compact@claude-modsMods install like plugins. Run /reload-plugins or start a new session to load it.
What it can access
- Reach
- L2Writes or runs: drives Claude, reads the prompt box, draws, other: $.decision.ask
- Sees
- every tool call, every prompt
- Hooks
session.starttool.callprompt.submitturn.complete
From a static scan of the mod's code by awesome-claude-code-mods. It shows what the code asks for, not a security review.
How Claude Code mods work
- A mod is a plugin with a JavaScript or TypeScript module that runs inside Claude Code (version 2.1.287 or later). It can draw panes beside the transcript, a band above the prompt, buttons and toasts, restyle the spinner and tool rows, and watch or rewrite tool calls and prompts.
- It installs like any plugin: add its marketplace with /plugin marketplace add owner/repo, then /plugin install name@marketplace.
- Check what a mod can reach before installing: it runs with your permissions. Reach goes from drawing only, to reading files, to writing files or running processes, to the network. Run claude plugin validate on a clone to list its hooks and calls without running it.
- Turn one off from the Installed tab in /plugin, or start Claude Code with --safe-mode to skip every customization for a session.
Details
- Repository
- kzarzycki/claude-mods
- Folder
plugins/self-compact- Marketplace
claude-mods- Type
- Workflow mods
- License
- NOASSERTION
- Last updated
- 28 October 2026
- Listed in
- awesome-claude-code-mods
FAQ
What is self-compact?
Compaction at task boundaries: the model can ask to compact once its turn ends. Above a fill threshold it is asked whether to, after a turn decision-model judges a finished task, or, without decision-model, with each prompt. It is a mod for Claude Code, from kzarzycki/claude-mods on GitHub.
What can the self-compact mod access?
Its widest reach is "writes or runs" (drives Claude, reads the prompt box, draws, other: $.decision.ask). It sees every tool call, every prompt. Like any mod, it runs with your permissions.
How do I install self-compact?
Inside Claude Code (v2.1.287 or later), run /plugin marketplace add kzarzycki/claude-mods and then /plugin install self-compact@claude-mods.
Is self-compact popular?
kzarzycki/claude-mods has 1 GitHub stars and 0 forks, and is included in 1 curated Claude Code list. It was last updated on 28 October 2026.
More workflow mods
rtl-text
1aliir74/claude-code-rtl
Shapes and reorders Persian, Arabic and Hebrew text in the transcript through fribidi so the terminal draws it joined, right-to-left and right-aligned.
semantic-grep
1alex4o/system-one-mods
A 'search' tool: ranks every function (or file) under a glob by a System One model (Jev, or the open-source decider) against a plain-language query.
rice-multi-models
1moabualruz/rice-multi-models
Codex (ChatGPT subscription), OpenCode Zen and Go, and configured API or local providers as session and subagent models, no proxy
stepwarden
1getexcited/stepwarden
Every tool call your agent makes, checked before it runs. Verifies each pending tool call against the session's plan using TypeSafe AI's Jev model, then allows it, asks you, or blocks it.
read-stub
1alex4o/system-one-mods
On a full Read of a big source file, a System One model (Jev, or the open-source decider) decides per function/method/const whether to include it; the rest become stubs.
switchboard-notch
1sameeeeeeep/switchboard-notch
Claude's questions drop from your Mac's notch as a native card. Pick with 1-4, type your own answer, or press esc to answer in Claude Code as usual.
Browse the most popular Claude Code mods, all workflow mods, or the skills and MCP server directories.