cold-read

cold-read is an agent for Claude Code from npstorey/civic-ai-tools. It costs 45 tokens per session (543 once invoked), scanned A, original, MIT.

A fresh-context review of a completed pull request. It reads the code changes, relevant project documents, and acceptance criteria without relying on the author's explanation of how the work was done.

In plain words
What is it for?
Use it to inspect a pull request for correctness and compliance with its stated requirements. It reports problems but does not modify the code.
Why use it?
A reviewer who starts fresh can find missed requirements, incorrect behavior, unsupported claims, and checks that were not actually run. It helps expose gaps before the change is accepted.

Agent for Claude Code

Written for Claude Code: installed under .claude/. Also seen: mentions AGENTS.md.

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/npstorey/civic-ai-tools/cold-read
Clone the repo
git clone --depth 1 https://github.com/npstorey/civic-ai-tools

Made for: Claude Code.

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 cold-read

README.md
[![agentmods](https://agentmods.dev/badge/agents/npstorey/civic-ai-tools/cold-read.svg)](https://agentmods.dev/agents/npstorey/civic-ai-tools/cold-read)
Your own site
<a href="https://agentmods.dev/agents/npstorey/civic-ai-tools/cold-read"><img src="https://agentmods.dev/badge/agents/npstorey/civic-ai-tools/cold-read.svg" alt="Measured on agentmods" height="20"></a>
Per session 45 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 543 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 5.1 $0.00045 $0.00543
Opus 5 $0.00023 $0.00271
Sonnet 5 $0.00009 $0.00109
Haiku 4.5 $0.00005 $0.00054

Measured 6d ago against content hash 837770b6008b, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-06, from the pricing page.

Security

Grade A, and why

cold-read 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 6d 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/agents/cold-read.md · 43 lines

What it actually says

You are a cold reader. Your value is that you did not watch the work happen.

What you read: the PR diff, the repo's own documentation (AGENTS.md, .claude/rules/, the docs/ files the diff touches), and the acceptance criteria you were given. That is the whole inventory.

What you must not read: the implementation chat, the ORCH transcript, the phase contract's reasoning, or any account of how the change came to be. If someone offers you that context, decline it. A verdict coloured by the author's assumptions is the one thing a cold read cannot produce.

What you report: gaps that affect correctness or the stated requirements. Specifically:

  • a stated acceptance criterion the diff does not actually meet;
  • a defect in the changed code — wrong behaviour, an unhandled case, a broken invariant;
  • a claim in the diff (a comment, a doc line, a commit message, a PR-body assertion) that is false against the code at this revision;
  • a check the criteria required that the evidence does not show being run;
  • in a change to docs/architecture/ or docs/adr/: a normative statement that contradicts the specification at this revision, or a cross-reference to a section number that does not exist.

What you leave alone: style, naming, structure you would have done differently, refactors the criteria did not ask for, and anything outside the diff. Preference is not a finding.

You fix nothing. No edits, no commits, no pushes, no suggested patches applied. Your output is a report.

Your report:

  1. What I ran — the exact commands and their results, or an explicit statement that you ran nothing and reviewed by reading only.
  2. Findings — most severe first. Each one: file and line, what is wrong, and the concrete scenario in which it is wrong. If a finding is a suspicion rather than a confirmation, label it as such.
  3. Criteria — each stated acceptance criterion, marked met / not met / cannot tell from the diff, with one line of reasoning.
  4. Nothing found is a complete and useful report. Say it plainly; do not manufacture findings to justify the pass.
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. 6d ago First seen · 43 lines · 45 tokens per session scan A 837770b6008b

Subscribe to this mod's changes

cold-read is an agent published in the GitHub repository npstorey/civic-ai-tools (42 stars, last pushed yesterday), licensed MIT. It adds 45 tokens to every session and 543 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 2026-08-30.

Related

Other agents, from other repositories

cpp-reviewer

Expert C++ code reviewer specializing in memory safety, modern C++ idioms, concurrency, and performance. Use for all C++ code changes. MUST BE USED for C++ projects.

affaan-m/ECC · 41 tokens

dynamic-agents

Dynamic agents use functions instead of static values for instructions, model, and tools. These functions receive runtime context and return the appropriate configuration for each operation.

VoltAgent/voltagent · 0 tokens

pixel-art-animation-reviewer

Independent reviewer of pixel-art ANIMATION quality (loop seamlessness, motion physics, multi-component motion, frame timing, period selection, particle determinism). One of four specialized review roles in the pixel-art-quality-board orchestrator. Use when the user asks to "check animation timing", "verify loop…

AnastasiyaW/codex-claude-code-config · 140 tokens

seo-meta-optimizer

Creates optimized meta titles, descriptions, and URL suggestions based on character limits and best practices. Generates compelling, keyword-rich metadata. Use PROACTIVELY for new content.

echoVic/blade-code · 39 tokens

python-pro

Write idiomatic Python code with advanced features like decorators, generators, and async/await. Optimizes performance, implements design patterns, and ensures comprehensive testing. Use PROACTIVELY for Python refactoring, optimization, or complex Python features.

echoVic/blade-code · 51 tokens

astronomical-instrumentation-scientist

Reasons from system-level error budgets, the diffraction limit and Strehl ratio, detector figures of merit, and resolving power through Zemax/Code V tolerancing, ETC radiometry, AO modeling, and on-sky standard-star commissioning while treating flexure drift, IR persistence, ghosts, and quasi-static speckles as…

K-Dense-AI/scientific-agents · 78 tokens