agentmods.dev

sonmat

A workflow plugin that checks an AI coding agent's work from several separate viewpoints, including safety, completeness, reasoning, and whether the result matches the request. It can also coordinate repeated checking and retain notes after work ends.

In plain words
What is it for?
Use it for code-task verification, second checks against the original request, finding missing work or unsafe files, challenging assumptions, and carrying useful notes into later tasks.
Why use it?
It reduces the chance that an agent will approve its own incomplete, unsafe, or misunderstood work. Separate checks can expose problems that the main working process missed.

Plugin for Claude Code

6 4d ago A scan tokens not measured original BSD-3-Clause
View the source file jun0-ds/sonmat
Install

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.

Claude Code
/plugin marketplace add jun0-ds/sonmat
agentmods
npx agentmods add plugins/jun0-ds/sonmat/marketplace
Clone the repo
git clone --depth 1 https://github.com/jun0-ds/sonmat

Made for: Claude Code.

Per session not measured What this adds to a session before it is invoked.
When invoked not measured Not applicable: nothing here is loaded into a session.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Security

Grade A, and why

sonmat 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 2d ago.

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.

.claude-plugin/marketplace.json · 32 lines

What it actually says

{
  "name": "sonmat",
  "owner": {
    "name": "jun0-ds"
  },
  "metadata": {
    "description": "sonmat — External verification system. Six verification axes (guard/inspect/witness/punch/devil/scribe) + autoloop orchestration. Discovery-led depth, protocol-isolated intent-artifact comparator (witness), post-work persistence layer (scribe).",
    "version": "0.17.1"
  },
  "plugins": [
    {
      "name": "sonmat",
      "source": "./",
      "description": "sonmat — External verification system. Six verification axes (guard/inspect/witness/punch/devil/scribe) + autoloop orchestration. Discovery-led depth, protocol-isolated intent-artifact comparator (witness), post-work persistence layer (scribe).",
      "version": "0.17.1",
      "author": {
        "name": "jun0-ds"
      },
      "keywords": [
        "verification",
        "autoloop",
        "autonomous",
        "discipline",
        "memory",
        "witness",
        "discovery-led"
      ],
      "category": "workflow"
    }
  ]
}
Files

What ships with it

1 file beside marketplace.json in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.

Changes

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.

  1. 2d ago First seen · 32 lines scan A e9a7fc32e500

Subscribe to this mod's changes

sonmat is a plugin published in the GitHub repository jun0-ds/sonmat (6 stars, last pushed 4d ago), licensed BSD-3-Clause. 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.

PlanningProductivityai-toolsai-verificationanthropicautonomous-agentclaudeclaude-code
Related

Other plugins, from other repositories

llm-wiki

LLM-powered personal wiki — autonomous knowledge base with research-on-miss, ingestion, search, a browsable web UI, and universal data gravity. Saves and retrieves knowledge automatically whenever relevant.

Oshayr/LLM-Wiki · not measured

llm-wiki

Plugin marketplace listing 1 plugin: llm-wiki.

Oshayr/LLM-Wiki · not measured

safe-android-reverser

Sandboxed Android reverse engineering with semantic program indexing, XREF/CFG queries and structured network evidence through MCP.

salingnh/android-reverse-engineering-mcp · not measured

unslop

Three anti-slop scanners in one bundle: a-unslop-text catches AI-diction and cadence tells in prose, a-unslop-code catches bugs and tutorial-shaped patterns in source, a-unslop-ui catches stock-default looks in web UI. Each ships a deterministic scanner plus a Build mode that heads off the tell before it ships. The…

imisic/claude-marketplace · not measured

lovstudio-cc-plugins

Plugin marketplace listing 1 plugin: lovstudio.

lovstudio/cc-plugins · not measured

swift-performance-pro

Performance: Instruments, SwiftUI render cost, memory and retain cycles, launch time, lists.

laxrajpurohit/swift-skills-pro · not measured