Plugin Claude Code
Safety guardrails, cross-host hooks, and agentdb memory for Claude Code and Codex in auto mode.
12 yesterday A
tokens not measured
original MIT
kernel
02Plugin Claude Code
KERNEL is a Claude Code and Codex plugin that replaces per-action approval prompts with guardrails enforced by hooks. Hook context is shaped for both hosts' plain and structured output parsers. In auto mode, destructive commands and secret writes are blocked before they run; irreversible operations require a one-time…
12 yesterday A
tokens not measured
original MIT
PreToolUse
03Hook Claude Code
Runs before the agent uses a tool for Bash tool calls, executing release-guard.sh. From ariaxhan/kernel-claude.
12 yesterday A
tokens not measured
copy · 55% MIT
Settings file Claude Code
Agent settings declaring 1 hook event (PreToolUse).
12 yesterday A
tokens not measured
original MIT
Instructions file CodexOpenCode
Instructions for ariaxhan/kernel-claude, a project described as: Claude Code learns from itself. Persistent memory, multi-agent orchestration, and a scientific experiment engine that proves which rules actually work.
12 yesterday A 6,558 tokens
original MIT
Instructions file
Instructions for ariaxhan/kernel-claude, a project described as: Claude Code learns from itself. Persistent memory, multi-agent orchestration, and a scientific experiment engine that proves which rules actually work.
12 yesterday A 6,566 tokens
copy · 94% MIT
Agent
Saved: 2026-03-12T19:31:15Z Trigger: auto compact Branch: main.
12 yesterday A 0 tokens
original MIT
Agent
Saved: 2026-03-12T19:48:00Z Trigger: auto compact Branch: main.
12 yesterday A 0 tokens
copy · 86% MIT
Agent
Saved: 2026-03-12T19:09:20Z Trigger: manual compact Branch: main.
12 yesterday A 0 tokens
copy · 89% MIT
adversary
10Agent
QA - assume broken, find edge cases, prove with evidence.
12 yesterday A 16 tokens
original MIT
Agent
Structurally separate eval agent. Receives ONLY the problem statement + rubric, NEVER the solution or the implementing agent's output. Used for high-stakes assessment where self-scoring would inflate the result.
12 yesterday A 45 tokens
original MIT
deep-diver
12Agent
Pre-implementation failure-mode research. Runs Research-Failures-First protocol: spawns Channel-A (GitHub issues) + Channel-D (production case studies) in parallel, merges into canonical failure-mode map at meta/research/ .md. Gates non-trivial native/infra/schema work.
12 yesterday A 66 tokens
original MIT
dreamer
13Agent
Multi-perspective debate agent. Generates minimalist, maximalist, and pragmatist approaches grounded in actual codebase context.
12 yesterday A 27 tokens
original MIT
lane-worker
14Agent
Isolated implementation lane for a commissioned parallel burn. Owns exactly one file-disjoint slice of a larger contract, follows the pilot/spec verbatim, checkpoints progress to AgentDB, and never commits. Spawned only when an orchestrator has genuinely independent, file-disjoint work to run concurrently.
12 yesterday A 62 tokens
original MIT
researcher
15Agent
Deep research agent. Finds existing solutions, evaluates packages, documents pitfalls before any implementation begins. Spawned by orchestrator when encountering unfamiliar tech, new integrations, or package selection decisions.
12 yesterday A 39 tokens
original MIT
reviewer
16Agent
Code review agent. Reviews PRs and code changes for quality.
12 yesterday A 16 tokens
original MIT
scout
17Agent
Codebase reconnaissance agent. Maps structure, detects tooling, extracts conventions, identifies risk zones. Spawned by orchestrator on first interaction with unfamiliar codebase or when discovery is needed.
12 yesterday A 39 tokens
original MIT
surgeon
18Agent
Minimal diff implementation, commit every working state.
12 yesterday A 12 tokens
original MIT
Agent
Forensic miner of session transcripts and git history. Reads large JSONL corpora (Claude and Codex session logs), extracts failure timelines, recurring patterns, and cited evidence, and returns the conclusion, not the raw pile. Read-only. Spawned to protect the main context from megabytes of logs when only the…
12 yesterday A 74 tokens
original MIT
SessionStart
20Hook
Runs when a session starts, executing session-start.sh, autopush.sh and knowledge-graph.sh (3 commands). From ariaxhan/kernel-claude.
12 yesterday A
tokens not measured
copy · 67% MIT
PostToolUse
21Hook
Runs after a tool call finishes for Bash, WebFetch, WebSearch, mcp__.*, Write and Edit tool calls, executing syntax-coach.py, scan-output.sh, log-write.sh and validate-json-schema.sh (4 commands). From ariaxhan/kernel-claude.
12 yesterday A
tokens not measured
original MIT
Hook
Runs when the agent asks permission for an action for Bash tool calls, executing auto-approve-safe.sh. From ariaxhan/kernel-claude.
12 yesterday A
tokens not measured
original MIT
Hook
Runs when a tool call fails for Edit, Write and Bash tool calls, executing capture-error.sh. From ariaxhan/kernel-claude.
12 yesterday A
tokens not measured
original MIT
Hook
Runs when you submit a prompt, before the agent sees it, executing context-usage.py, route-request.sh and post-compact-restore.sh (3 commands). From ariaxhan/kernel-claude.
12 yesterday A
tokens not measured
original MIT