scorekeeper

A tracking system for recording what an AI agent has committed to say or do. It checks whether the agent stays within those commitments.

In plain words
What is it for?
Use it to monitor an agent’s promises, claims, and permissions during a task. It is intended for controlling agent memory and limiting overreach.
Why use it?
It helps catch when an agent changes its story, claims more than it established, or acts beyond the agreed scope.

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 michalstrnadel/scorekeeper
agentmods
npx agentmods add plugins/michalstrnadel/scorekeeper/claude-code-plugin
Clone the repo
git clone --depth 1 https://github.com/michalstrnadel/scorekeeper

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

scorekeeper 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-code-plugin/.claude-plugin/plugin.json · 14 lines

What it actually says

{
  "name": "scorekeeper",
  "description": "Commitment tracking for LLM agents — a deontic scoreboard of what the agent is entitled to claim and to do; catches drift and overreach.",
  "version": "0.4.0",
  "author": {
    "name": "Michal Strnadel",
    "url": "https://github.com/michalstrnadel/scorekeeper"
  },
  "homepage": "https://github.com/michalstrnadel/scorekeeper",
  "repository": "https://github.com/michalstrnadel/scorekeeper",
  "license": "Apache-2.0",
  "keywords": ["commitment-tracking", "agent-memory", "context-engineering", "inferentialism", "scope-control", "overreach"]
}
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 · 14 lines scan A 01c8b1b9901a

Subscribe to this mod's changes

scorekeeper is a plugin published in the GitHub repository michalstrnadel/scorekeeper (2 stars, last pushed 1mo 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.

Related

Other plugins, from other repositories

aria-knowledge

ARIA — Applied Reasoning and Insight Architecture. Persistent human-governed knowledge for Claude Code via a five-phase lifecycle: capture → govern → promote → apply → refresh. Stages session insights/decisions/feedback into review backlogs, promotes approved items to a tag-indexed markdown base, and applies them via…

mikeprasad/aria-knowledge · not measured

aria-knowledge

ARIA — Applied Reasoning and Insight Architecture. Knowledge management, decision discipline, codebase understanding, plan pre-mortems, and release retrospectives for Claude Code.

mikeprasad/aria-knowledge · not measured

aria-cowork

Knowledge management for Claude Cowork — capture, organize, and apply durable knowledge across your work, available in any project workspace via a persistent folder grant. Companion to aria-knowledge for Claude Code; both share one knowledge folder so insights flow between conversational and engineering work. Default…

mikeprasad/aria-knowledge · not measured

aria-probe

Validation probe for aria-cowork. Tests whether a Cowork plugin can read and write the user-attached knowledge folder, verifies bidirectional access with aria-knowledge in Code, and (v0.2.0) tests whether Cowork can read the existing aria-knowledge config at /.claude/aria-knowledge.local.md when /.claude/ is attached.

mikeprasad/aria-knowledge · not measured

claude-code

HaloMem organizational memory: session-start recall, session-end capture, MCP tools. Configure with HALOMEMURL and HALOMEMAPIKEY.

halofyai/halofy · not measured

titen-memory

Use Titen's evidence-grounded collaborative memory. The local stdio mode needs no configuration; a served Titen endpoint is the alternative for a shared store.

RamaAditya49/titen · not measured