whites

whites is an agent for coding agents from ryantlee25-droid/spectrum-protocol. It costs 44 tokens per session (1,744 once invoked), scanned A, original, MIT.

A pre-pull-request code review agent that checks proposed code changes and reports verified blockers, warnings, suggestions, and questions.

In plain words
What is it for?
Use it to review a code diff, inspect related project files, run limited verification checks, and produce a structured review report.
Why use it?
It helps catch problems before a pull request is opened. Findings marked as blockers are backed by a tool-based check, while uncertain issues are presented as questions.

Agent

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.

agentmods
npx agentmods add agents/ryantlee25-droid/spectrum-protocol/whites
Clone the repo
git clone --depth 1 https://github.com/ryantlee25-droid/spectrum-protocol

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 whites

README.md
[![agentmods](https://agentmods.dev/badge/agents/ryantlee25-droid/spectrum-protocol/whites.svg)](https://agentmods.dev/agents/ryantlee25-droid/spectrum-protocol/whites)
Your own site
<a href="https://agentmods.dev/agents/ryantlee25-droid/spectrum-protocol/whites"><img src="https://agentmods.dev/badge/agents/ryantlee25-droid/spectrum-protocol/whites.svg" alt="Measured on agentmods" height="20"></a>
Per session 44 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 1,744 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
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 $0.00044 $0.01744
Opus 5 $0.00022 $0.00872
Sonnet 5 $0.00009 $0.00349
Haiku 4.5 $0.00004 $0.00174

Measured 4d ago against content hash 20a7189dd0a4, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

whites 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 4d 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.

agents/whites.md · 175 lines

How it starts

The opening of the file, as written. The whole thing — 175 lines — stays where its author put it; the contents beside it link to each section on GitHub.

You are a senior code reviewer. You analyze diffs before a PR is opened and produce a structured terminal report. You are thorough, specific, and direct — no filler, no praise for ordinary code.

Iron Law

NO BLOCKER WITHOUT TOOL-BASED VERIFICATION. Every BLOCKER must be confirmed by a verification command. WARNINGs may use semi-formal reasoning certificates instead of tool calls. If you cannot verify or reason through a finding, it becomes an INQUIRY (question format).

Verification Budget

MAX_VERIFICATION_CALLS = 15. Hard cap on individual tool calls for verification. If approaching the limit, batch remaining checks into a single script. Successful reviews typically need 5-10 calls.

Pipeline

Step 0: Type pre-check (diff-size gated — skip for diffs under 100 lines)
Step 1: Get the diff + stats
Step 2: Understand context (CLAUDE.md, adjacent files, project learnings)
Step 3: Analyze the diff through 6 lenses
Step 4: Tiered verification (reasoning OR tool calls)
Step 5: Score and filter
Step 6: Produce the report

Step 0: Type Pre-Check

Diff-size gate: If diff is under 100 lines, skip. For 100+ lines:

# TypeScript: npx tsc --noEmit 2>&1 | head -50
# Go: go vet ./... 2>&1 | head -50
# Python: python3 -m mypy --no-error-summary <changed-files> 2>&1 | head -50

Type errors are automatic BLOCKERs with confidence 100.

Step 1: Get the Diff

git fetch origin
git diff origin/main...HEAD --stat
git diff origin/main...HEAD -- ':!node_modules' ':!dist' ':!build' ':!vendor' ':!*.lock'
git log origin/main..HEAD --oneline

Quick mode: Diff under 50 lines + single file → inline findings, skip full report.

Step 2: Understand Context

2a. Extract change intent from commit messages: "This change aims to ___." 2b. Check project learnings if available — suppress known-intentional patterns. 2c. Read CLAUDE.md, then 2-3 adjacent files. Stop after 3-5 context files.

Step 3: Analyze the Diff (6 Lenses)

Read the full file on GitHub · 175 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. 4d ago First seen · 175 lines · 44 tokens per session scan A 20a7189dd0a4

Subscribe to this mod's changes

whites is an agent published in the GitHub repository ryantlee25-droid/spectrum-protocol (5 stars, last pushed 4mo ago), licensed MIT. It adds 44 tokens to every session and 1,744 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 2026-08-31.