hunch

An engineering-memory and change-review system for AI-assisted codebases. It records design decisions and rules, then checks proposed changes against those rules and their possible effects on other code.

In plain words
What is it for?
Recording architectural decisions and constraints, explaining why code is structured a certain way, checking change impact, and deciding whether a change is safe to merge.
Why use it?
It helps developers retain the reasoning behind a codebase and catch changes that break established rules. This reduces dependence on undocumented knowledge about how the system should behave.

Plugin for Claude Code

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 davesheffer/hunch
agentmods
npx agentmods add plugins/davesheffer/hunch/marketplace
Clone the repo
git clone --depth 1 https://github.com/davesheffer/hunch

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

hunch 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 · 24 lines

What it actually says

{
  "name": "hunch",
  "owner": {
    "name": "Dave Sheffer",
    "email": "[email protected]"
  },
  "description": "Hunch — engineering memory and a deterministic Change Gate for AI-assisted codebases.",
  "plugins": [
    {
      "name": "hunch",
      "source": "./plugin",
      "description": "The hunch_* MCP tools (why, constraints, blast radius, merge verdict, capture) plus /hunch:capture, /hunch:heal, /hunch:why, /hunch:fix, /hunch:fragile. Requires Node 22.13+; launches the exact npm package `@davesheffer/[email protected]` per project.",
      "author": {
        "name": "Dave Sheffer",
        "email": "[email protected]"
      },
      "homepage": "https://hunch-pi.vercel.app",
      "repository": "https://github.com/davesheffer/hunch",
      "license": "Apache-2.0",
      "keywords": ["engineering-memory", "architectural-conformance", "decisions", "invariants", "mcp"]
    }
  ]
}
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 · 24 lines scan A a76ed6bda868

Subscribe to this mod's changes

hunch is a plugin published in the GitHub repository davesheffer/hunch (6 stars, last pushed 2d ago), licensed Apache-2.0. 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.