codebase-grounding-judge

codebase-grounding-judge is an agent for Claude Code from closedloop-ai/claude-plugins. It costs 42 tokens per session (1,072 once invoked), scanned A, original, Apache-2.0.

An automated reviewer that checks whether an implementation plan matches evidence gathered from the real codebase. It looks for invented file paths, modules, functions, APIs, and integration points.

In plain words
What is it for?
Use it to review file paths, module names, imports, existing-code claims, and proposed integration points in a plan.
Why use it?
It helps catch plans that sound plausible but refer to code that does not exist. This reduces wasted implementation work caused by planning without enough codebase context.

Agent for Claude Code

Written for Claude Code: shipped in a Claude Code plugin. Also seen: model in frontmatter.

Part of the judges plugin — 3 skills, 22 agents shipped together

Good fit Use it to review file paths, module names, imports, existing-code claims, and proposed integration points in a plan.

Compare 6 agents from other repositories ↓
Install with agentmods
npx agentmods add agents/closedloop-ai/claude-plugins/codebase-grounding-judge
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.

Clone the repo
git clone --depth 1 https://github.com/closedloop-ai/claude-plugins

Made for: Claude Code.

Or install judges, the plugin that ships this one along with the rest of its 3 skills, 22 agents.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

agentmods badge for codebase-grounding-judge

README.md
[![agentmods](https://agentmods.dev/badge/agents/closedloop-ai/claude-plugins/codebase-grounding-judge.svg)](https://agentmods.dev/agents/closedloop-ai/claude-plugins/codebase-grounding-judge)
Your own site
<a href="https://agentmods.dev/agents/closedloop-ai/claude-plugins/codebase-grounding-judge"><img src="https://agentmods.dev/badge/agents/closedloop-ai/claude-plugins/codebase-grounding-judge.svg" alt="Measured on agentmods" height="20"></a>
Per session 42 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 1,072 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5.1 $0.00042 $0.01072
Opus 5 $0.00021 $0.00536
Sonnet 5 $0.00008 $0.00214
Haiku 4.5 $0.00004 $0.00107

Measured today against content hash 275f8a6066d3, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-08, from the pricing page.

Security

Grade A, and why

codebase-grounding-judge 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 today.

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.

plugins/judges/agents/codebase-grounding-judge.md · 109 lines

How it starts

The opening of the file, as written. The whole thing — 109 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Codebase Grounding Judge

You are evaluating whether an implementation plan accurately reflects the real codebase — or whether it hallucinates structure, files, modules, or APIs that don't exist.

This judge is the primary differentiator between plans written with codebase context (CL) and plans written blind (OOTB).

Input Files

Read from $SYMPHONY_WORKDIR:

  1. investigation-log.md — ground truth about the codebase: real files, modules, conventions, patterns
  2. plan.json — the plan to evaluate
  3. prd.md — the requirements (for context on what's genuinely new)

If investigation-log.md is absent: Score all metrics 0.5 with justification "No investigation log — claims unverifiable." Do not attempt to evaluate. Do not access the filesystem.

What to Evaluate

Extract from plan.json:

  • File paths proposed in pendingTasks[].file
  • Module/class/function names referenced in task descriptions and content
  • Import statements and dependency references
  • Claims about what "already exists" in the codebase
  • Proposed integration points with existing code

Then cross-reference each claim against investigation-log.md:

Metric 1: file_path_accuracy (threshold: 0.8)

For each file, class, method, database table, variable and other constructs in the plan:

  • Existing file being modified: Is it confirmed in investigation-log? If yes, correct.
  • New file being created: Is its directory consistent with project structure shown in investigation-log? If yes, plausible.
  • Hallucinated path: Path contradicts investigation-log structure (wrong directory depth, wrong naming convention, module doesn't match project layout). Flag as hallucination.

Score:

  • 1.0: All paths either confirmed by investigation-log or structurally consistent with it
  • 0.5: 1–2 paths are unverifiable or inconsistent but no clear contradictions
  • 0.0: One or more paths directly contradict investigation-log (e.g., file claimed to exist that doesn't, path uses wrong module structure)

Read the full file on GitHub · 109 lines

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. today First seen · 109 lines · 42 tokens per session scan A 275f8a6066d3

Subscribe to this mod's changes

codebase-grounding-judge is an agent published in the GitHub repository closedloop-ai/claude-plugins (103 stars, last pushed yesterday), licensed Apache-2.0. It adds 42 tokens to every session and 1,072 once invoked, about $0.0002 per session on Opus 5. 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 2026-09-07.

Related

Other agents, from other repositories

flutter-reviewer

Flutter and Dart code reviewer. Reviews Flutter code for widget best practices, state management patterns, Dart idioms, performance pitfalls, accessibility, and clean architecture violations. Library-agnostic — works with any state management solution and tooling.

affaan-m/ECC · 50 tokens

vue-reviewer

Expert Vue.js code reviewer specializing in Composition API correctness, reactivity pitfalls, component architecture, template security, and Vue-specific performance. Use for any change touching .vue, .ts/.js files with Vue imports, or Vue ecosystem code (Pinia, Vue Router, Nuxt). MUST BE USED for Vue projects.

affaan-m/ECC · 68 tokens

harmonyos-app-resolver

HarmonyOS application development expert specializing in ArkTS and ArkUI. Reviews code for V2 state management compliance, Navigation routing patterns, API usage, and performance best practices. Use for HarmonyOS/OpenHarmony projects.

affaan-m/ECC · 49 tokens

csharp-reviewer

Expert C# code reviewer specializing in .NET conventions, async patterns, security, nullable reference types, and performance. Use for all C# code changes. MUST BE USED for C# projects.

affaan-m/ECC · 44 tokens

php-reviewer

Expert PHP code reviewer specializing in PSR-12 compliance, PHP type system, Eloquent ORM patterns, security, and performance. Use for all PHP code changes. MUST BE USED for PHP projects.

affaan-m/ECC · 44 tokens

fastapi-reviewer

Reviews FastAPI applications for async correctness, dependency injection, Pydantic schemas, security, OpenAPI quality, testing, and production readiness.

affaan-m/ECC · 33 tokens