Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
/plugin marketplace add kelp/agent-pluginsnpx agentmods add plugins/kelp/agent-plugins/knowledge-forgegit clone --depth 1 https://github.com/kelp/agent-pluginsGrade A, and why
knowledge-forge scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured yesterday.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
What it actually says
{
"name": "knowledge-forge",
"description": "Cross-session routing and capture for a three-layer personal knowledge base. /kb-research-policy teaches the index-first retrieval policy. /kb-capture files notes into the right wiki bucket with correct frontmatter.",
"version": "0.2.0",
"author": {
"name": "Travis Cole"
},
"homepage": "https://github.com/kelp/agent-plugins",
"repository": "https://github.com/kelp/agent-plugins"
}
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- yesterday First seen · 11 lines scan A 8e6785433b7c
knowledge-forge is a plugin published in the GitHub repository kelp/agent-plugins (2 stars, last pushed 8d ago), licensed MIT. Its token cost is not measured: this kind of file is read by the harness, not the model. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other plugins, from other repositories
zvm
Install, pin, and switch Zig toolchains with zvm (Zig Version Manager). Teaches agents zvm's real command surface and how to select a Zig version without disturbing the user's global install.
learnable
Learn any subject properly, in courses you actually finish. Four commands: learn is the way in and the way back, new-course researches a whole domain with a fleet of agents and turns it into a taxonomy and a plan aimed at your real gaps, self-assess measures your true level with a scenario quiz that cannot be gamed…
open-design-guide
Teaches Claude to use the Open Design content library: 151 design systems with real tokens, 71 design and frontend skills, 114 rendering templates. Installs the library as a sparse clone, no desktop app and no daemon.
hello-world
A simple example plugin that demonstrates basic Claude Code plugin functionality.
cs2-modding
Teach your coding agent to write Cities: Skylines II code mods (Claude Code, Codex CLI): the game's architecture, the modding techniques and the game mechanics behind them, every claim verified against the decompiled game. Knowledge only, no MCP server.
study-with-claude-code
AI 驱动的通用系统化学习 + 复习系统：roadmap 级知识树（评审子智能体查漏补缺、可增量更新）、递进式教学教案、苏格拉底式闭环学习 + 复习、独立子智能体评估掌握度、按艾宾浩斯曲线安排复习。A general-purpose systematic learning + review system as Claude Code skills.