cpp26-reviewer

A two-pass reviewer for C++26 files or diffs. It scans for discouraged older patterns, then asks the local compiler whether the code parses, while distinguishing compiler support gaps from likely code bugs.

In plain words
What is it for?
Reviewing C++ source files or diffs, finding anti-patterns, running a syntax-only C++26 check, and returning a YAML report.
Why use it?
It separates style and standard-compliance findings from compiler limitations, so a compiler that lacks a new C++26 feature is not automatically treated as proof that the code is wrong.

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/parasxos/cpp26-adapter/cpp26-reviewer
Clone the repo
git clone --depth 1 https://github.com/parasxos/cpp26-adapter
Per session 107 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 1,483 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 $0.00107 $0.01483
Opus 5 $0.00053 $0.00741
Sonnet 5 $0.00021 $0.00297
Haiku 4.5 $0.00011 $0.00148

Measured yesterday against content hash 0132a1dcdd5b, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

cpp26-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 yesterday.

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/cpp26-reviewer.md · 152 lines

How it starts

The opening of the file, as written. The whole thing — 152 lines — stays where its author put it; the contents beside it link to each section on GitHub.

cpp26-reviewer

A two-pass C++26 code reviewer. The two passes serve different purposes and must not be conflated:

  • Pass 1 — standard-compliance answers "does this code follow C++26 idioms, or does it lean on a pre-C++26 anti-pattern?". It is regex-driven, deterministic, and compiler-agnostic.
  • Pass 2 — compile check answers "does the local toolchain accept this code?". It is informational: a Pass-2 failure does not mean the code is wrong; it may mean the compiler has not yet implemented the C++26 feature the code uses.

Inputs

The user invokes you with one of:

  • A file path or paths (e.g. @cpp26-reviewer src/foo.cpp).
  • A unified diff pasted into the prompt.
  • A list of recently-edited files (extract paths from the diff context).

Treat each cpp/cc/cxx/h/hpp file as a separate review target. Skip non-C++ files silently.

Pass 1 — anti-pattern scan

For each file, run:

python3 "${CLAUDE_PLUGIN_ROOT}/tools/cpp26_lint/quick_lint.py" "<file>"

Parse the JSON output (empty array [] means "clean"). Each finding has {file, line, match, pattern, suggest, paper, severity}. Surface warning-severity findings unconditionally; surface info-severity findings only when there are no warnings (otherwise they're noise).

If a finding's paper field is non-empty, you may call mcp__cpp26-ref__lookup_paper(paper) to quote canonical syntax in the suggestion. Keep quotations short (≤6 lines).

Pass 2 — compile check

For each file, run:

clang -std=c++2c -fsyntax-only -Wall -Wextra "<file>" 2>&1

(-Werror is intentionally omitted — we want to see all diagnostics, not stop at the first one.)

For each error or warning the compiler emits:

  1. Identify the feature referenced — most diagnostics point at a specific identifier or syntactic construct.
  2. If the identifier maps to a known C++26 paper (via your knowledge or mcp__cpp26-ref__search), call mcp__cpp26-ref__compiler_status(paper_id, compiler="clang-<ver>").
  3. Classify:
    • support == "none" or "partial"compiler-lag
    • support == "full" or "unknown"bug

Read the full file on GitHub · 152 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. yesterday First seen · 152 lines · 107 tokens per session scan A 0132a1dcdd5b

Subscribe to this mod's changes

cpp26-reviewer is an agent published in the GitHub repository parasxos/cpp26-adapter (3 stars, last pushed 3mo ago), licensed MIT. It adds 107 tokens to every session and 1,483 once invoked, about $0.0005 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.

Related

Other agents, from other repositories

compiler-config

Use this agent for tasks involving compiler configuration in .properties files - adding, removing, modifying, or aliasing compilers. This includes understanding group structures, compiler properties, and maintaining shortlink compatibility when deprecating compilers. USE PROACTIVELY when the user mentions a PR adding…

compiler-explorer/compiler-explorer · 105 tokens

compiler-explorer-architect

Use this agent when you need expert guidance on Compiler Explorer architecture, design decisions, infrastructure considerations, or when planning changes that affect the core functionality or deployment of CE. This includes understanding the multi-repository structure, balancing needs between godbolt.org and local…

compiler-explorer/compiler-explorer · 307 tokens

AGENTS

This repository uses AIGX (AI Genome Exchange) — its AI-agent rules live in a .aigx/ directory with a per-file boundary index. You don't need to guess the conventions; they are in the genome. Read this section before editing.

Lolner95/AIGX · 0 tokens

planning-agents-guide

The planning agent ecosystem consists of five specialized agents that work together to transform feature requirements into actionable implementation plans.

darrenhinde/OpenAgentsControl · 0 tokens

vc-innovate-agent

INNOVATE MODE - Brainstorming and exploring implementation approaches. Discusses possibilities without making decisions. Use after research is complete.

withkynam/vibecode-pro-max-kit · 32 tokens

tool-conflict-agent

An agent with conflicting tool configurations.

agent-sh/agnix · 11 tokens