claude

nebula

Local, offline shield: flags fabricated confession, decision handback, claim beyond evidence and forced repetition in Claude's replies. No network, no telemetry, no training on your data.

0 on GitHubby Hypernymv0.4.20UNLICENSEDupdated 23 days ago
/

Plugin nebula from marketplace hypernym-nebula in HypernymAI/hypernym_nebula. Commands read from the repo's manifest — 22 days ago.

This plugin runs code on your machine — its hooks fire automatically on events like tool use and it starts local MCP servers. The exact commands are listed below; read them before installing.

What it installs

Skills (2)

  • hypernym-nebula

    When the user says "hypernym nebula", switch on the Nebula shield's soliton mode — a reply that peaks orange gets a vivid SHIELDED banner and is rephrased once. "hypernym nebula off" switches it off (pass "off" as the argument).

  • shield-reframe

    Use whenever a SHIELD (local) note or block names a behavior in the current turn — forced repetition, claim beyond evidence, decision handback, fabricated confession, unauthorized action — to reframe the reply before the pilot sees it. Also use on your own if you notice one of these in a draft.

Slash commands (11)

  • /nebula:shield-backfill

    Opt in — read ALL your past Claude Code sessions on this machine into the local nebula record (never uploaded). Runs at low priority.

  • /nebula:shield-exclude

    Never read, score or store sessions from matching projects (glob on the project folder name or working directory).

  • /nebula:shield-export

    Export everything nebula has stored to a JSON file in ~/.nebula (your data, readable).

  • /nebula:shield-include

    Remove a project exclusion.

  • /nebula:shield-privacy

    nebula privacy report — exactly what is stored on this machine, where, how much, and what is excluded. Nothing leaves your machine.

  • /nebula:shield-purge

    Delete everything nebula has stored (database, caches, logs). Without "yes" it only shows what would be deleted.

  • /nebula:shield-retention

    Choose what transcript text nebula keeps — none (default: scores and counts only), excerpt (first 300 chars), or full.

  • /nebula:shield-soliton

    Soliton — when a reply peaks orange on the shield, flag it with a vivid banner and have Claude rephrase it once. "on", "off", or nothing to see the current state.

  • /nebula:shield-status

    nebula status — on/off, mode, detectors and thresholds, what it has flagged recently, per model.

  • /nebula:shield-wings

    Choose the wings beside HYPERNYM.ai on the nebula badge — opens the numbered contact sheet; give a number to switch (e.g. 91), or on/off for all badge icons.

  • /nebula:statusline

    The hypernym-nebula status line (shield badge + Hypernym logo, and the Zephyr line when the hypernym plugin is installed). Set up automatically on session start; "remove" takes it out, "force" replaces someone else's status line (backed up first).

MCP servers (1)

  • nebula

    runs: sh -c r="${CLAUDE_PLUGIN_ROOT}"; s=bin/nebula-run; if [ ! -x "$r/$s" ]; then r=$(ls -td "$HOME"/.codex/plugins/cache/*/nebula/*/ 2>/dev/null | head -1); fi; exec "${r%/}/$s" mcp_server.py

Hooks (3)

  • SessionStart

    "${CLAUDE_PLUGIN_ROOT}/bin/nebula-run" session_start.py

  • UserPromptSubmit

    "${CLAUDE_PLUGIN_ROOT}/bin/nebula-run" hook_user_prompt.py

  • Stop

    "${CLAUDE_PLUGIN_ROOT}/bin/nebula-run" hook_stop.py

View source on GitHub hypernym.ai

Similar plugins

career-ops

claudecodex

Open-source AI job search that runs locally in your AI coding CLI: scan job portals, evaluate listings into a structured A-H report with a 1-5 score, tailor your CV, track applications

5 skills · hooks

career-ops-hq/career-ops 82,253

Agent runtimes — local WASM-sandboxed agents (rvagent) + Anthropic Claude Managed Agents (cloud); one interface, local-vs-cloud backends

5 skills · 2 commands · 9 agents

ruvnet/ruflo 81,589

pake

claudecodex

Package any website or local web build into a lightweight desktop app with Pake CLI.

1 skill

tw93/Pake 63,463

Use Alibaba Open Code Review from local Codex.

2 skills

alibaba/open-code-review 50,068

ifixai

claudecodex

Independent auditing for AI agents, open source. Claude discovers your setup, confirms the cost, and audits your agent against the one question other tools skip: is it doing the job it's supposed to do, given your business rules and org structure? Test any model (Anthropic, OpenAI, Gemini, Azure, Bedrock, …) graded by self, one independent judge, or a cross-vendor panel; each provider is billed to its own account via a key in your settings, never the command line. Results render as an interactive artifact (with a static-report fallback).

1 skill · hooks

ifixai-ai/iFixAi 45,158

Maintain nebula?

This page updates itself from your repo. Add the badge to your README, or get it Featured to pin it above the organic results with a followed link to your site.

claude & codex pluginslisted