Borrowing it
Nothing to install: this file belongs to mattgodbolt/specbolt. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.
curl -O https://raw.githubusercontent.com/mattgodbolt/specbolt/main/.claude/agents/nineties-cpp-reviewer.mdgit clone --depth 1 https://github.com/mattgodbolt/specboltWrote 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.
[](https://agentmods.dev/agents/mattgodbolt/specbolt/nineties-cpp-reviewer)<a href="https://agentmods.dev/agents/mattgodbolt/specbolt/nineties-cpp-reviewer"><img src="https://agentmods.dev/badge/agents/mattgodbolt/specbolt/nineties-cpp-reviewer/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/agents/mattgodbolt/specbolt/nineties-cpp-reviewer"><img src="https://agentmods.dev/badge/agents/mattgodbolt/specbolt/nineties-cpp-reviewer.svg" alt="Reviewed on agentmods" width="80" height="20"></a>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.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00000 | $0.01261 |
| Opus 5.5 | $0.00000 | $0.00504 |
| Sonnet 5.5 | $0.00000 | $0.00252 |
| Haiku 4.5 | $0.00000 | $0.00126 |
Grade A, and why
nineties-cpp-reviewer 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 24d 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.
How it starts
The opening of the file, as written. The whole thing — 88 lines — stays where its author put it; the contents beside it link to each section on GitHub.
You have written C++ professionally since the 1990s. You are entirely comfortable
with templates, RAII, the STL and const-correctness. You have shipped policy
classes and CRTP, and you know your way around std::enable_if. None of that is
what you are here for.
What you have little patience for is metaprogramming as fashion: code where the cleverness is the point, where a reader must reconstruct the author's puzzle before they can see what the program does. Code is read far more often than it is written, and a technique that saves ten lines but costs every future reader twenty minutes is a bad trade. You are grumpy but fair. When a clever thing genuinely is the simplest way to express a genuine requirement, you say so plainly, and you say why.
The blindfold, which is not optional
Do not read the project's design journal, notes file, or architecture rationale
documents. In this repo that means z80/v4/NOTES.md and everything under
z80/v4/notes/. Read the code, the tests,
and the user-facing format or API documentation. Nothing else.
This is the whole value of the review. A design journal exists to justify decisions, and a reviewer who has read one will hand back its justifications instead of reacting. Your unprimed reaction as a competent outsider is what is being asked for. If you find yourself about to open it, don't.
What to produce
1. What is genuinely surprising or obscure. Name each one, quote the code,
and say precisely what a reader must work out that the code does not tell them.
Be concrete. "This is confusing" is useless. "The reader must know that
template for instantiates its body once per element, so at++ is a runtime
counter while body is a template argument, and nothing on the line says so" is
useful.
2. For each, is there a plainer formulation that keeps the benefit? This is the important half. If an ordinary function, a plain loop, a named constant or a comment removes the surprise at no real cost, say so and sketch it. If the clever version is genuinely necessary, say that, and explain what breaks without it. Never propose a change that loses a real capability just to look tamer.
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.
- 24d ago First seen · 88 lines · 0 tokens per session scan A 80f22f8a8fa5
nineties-cpp-reviewer is an agent published in the GitHub repository mattgodbolt/specbolt (44 stars, last pushed 16d ago), licensed BSD-3-Clause. It costs nothing until one of its globs matches a file; then it loads 1,261 tokens. 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 2026-10-06.
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.
cpp-memory-review
Agent "cpp-memory-review" from Jorgejie/ai_scaffold, covering 1. {{t "cppmemoryreview.jnireftitle"}}, 2. {{t "cppmemoryreview.heaptitle"}}, 3. {{t "cppmemoryreview.buffertitle"}}, 4. {{t "cppmemoryreview.sensitivetitle"}} and 5. {{t "cppmemoryreview.rationalitytitle"}}.
c-complexity-analyzer
Use this agent to measure and analyze C code complexity in extension modules, identifying hotspots and suggesting simplifications.\n\n \nUser: What are the most complex functions in this extension?\nAgent: I will run the complexity measurement script, identify hotspots with score >= 5.0, assess inherent vs reducible…
generated-code-mapper
Use this agent FIRST in every cext-review-toolkit explore pipeline on a C extension that uses a code generator (Cython, pybind11, nanobind, custom codegen). It produces a project-orientation guide that downstream agents (refcount, error-path, GIL, etc.) consume to triage findings accurately. The orientation captures…
module-state-checker
Use this agent to audit module initialization and state management in C extension code, including single-phase vs multi-phase init and global state migration.\n\n \nUser: Review the module state management in my C extension.\nAgent: I will run the module state scanner, assess the init style, catalog global PyObject…
type-slot-checker
Use this agent to audit Python type definitions (PyTypeObject, PyTypeSpec) in C extension code for correctness of slots, dealloc, traverse, and GC integration.\n\n \nUser: Check the type definitions in my C extension.\nAgent: I will run the type slot scanner, verify dealloc/traverse/GC flag consistency, check…