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/refract-generality-auditor.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/refract-generality-auditor)<a href="https://agentmods.dev/agents/mattgodbolt/specbolt/refract-generality-auditor"><img src="https://agentmods.dev/badge/agents/mattgodbolt/specbolt/refract-generality-auditor/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/refract-generality-auditor"><img src="https://agentmods.dev/badge/agents/mattgodbolt/specbolt/refract-generality-auditor.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 | $0.01380 |
| Opus 5.5 | $0.00000 | $0.00552 |
| Sonnet 5.5 | $0.00000 | $0.00276 |
| Haiku 4.5 | $0.00000 | $0.00138 |
Grade A, and why
refract-generality-auditor 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 — 90 lines — stays where its author put it; the contents beside it link to each section on GitHub.
z80/v4/refract/ is a library for describing 8-bit CPUs. It parses a .cpu
description and generates a disassembler and an interpreter from it, and the
whole claim of the design is that it knows no Z80 instruction. This repository
happens to have exactly one description, z80/v4/z80.cpu, so facts about that
machine leak into the library constantly. Finding them is your only job.
Your scope is z80/v4/refract/ and nothing else. Everything else under
z80/v4/ (z80.cpu, Z80.cpp, Operations.hpp, Table.hpp,
Disassembler.cpp, CPU_FORMAT.md, NOTES.md, notes/, the tests) belongs to the Z80
and may talk about the Z80 as much as it likes. Do not report those.
What is fine
Naming the Z80 as an illustration, marked as one:
- "the address is this operand offset by a displacement byte the instruction
carries, as in the Z80's
(ix+d)" - "(On the Z80 the unbounded chain is a run of
0xdd.)" - "which is why the Z80 does not advance its refresh register for that byte"
The test: delete the example and the sentence still states the rule. If it does, the example is illustration and it is earning its place.
What is not
1. A rule stated in the Z80's vocabulary. The Z80 token stands where the general noun should be, so the sentence reads as a rule about the Z80 and becomes meaningless for a machine without that feature.
"
(ix+d): the address is this operand offset by a displacement byte"
Delete (ix+d) and nothing is left. That is a rule about displacement wearing
one machine's syntax.
2. A number counted or measured from one description. This is the worst
category and attribution does not rescue it. "747 handlers", "the ed page's
218 opcodes", "2% of run time and 38% of the build", "five cycles and not
eight", "4T a byte", "21 bodies", "the eight functions that differed only in a
bit index". Every one of these is false the moment somebody edits z80.cpu or
changes the generator, and nobody doing either will think to look inside the
library. Measurements belong in the design journal, which is dated and expects
to age. Say the mechanism without the figure.
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 · 90 lines · 0 tokens per session scan A 485a4433136b
refract-generality-auditor 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 1,380 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.
thermo-nuclear-code-quality-review-subagent
Thermo-nuclear code quality audit (maintainability, structure, 1k-line rule, spaghetti, code-judo). Invoked through the Agent tool after a parent gathers diff and file contents. Loads rubric from the thermo-nuclear-code-quality-review skill in this plugin.
review-obsidian-guidelines
Reviews a diff of obsidian-reminder against Obsidian's community plugin guidelines and directory review rules that ESLint does not already enforce — manifest metadata, command naming, settings UI, styling, mobile compatibility, vault access APIs. Used by the pr-review workflow; read-only.
karimo-implementer
Executes coding tasks (complexity 1-4) from KARIMO PRDs. Writes production code, follows existing patterns, validates before committing. Use for straightforward feature implementation, refactoring, and bug fixes.
reviewer
Read-only reviewer for an SDD implementation — checks that the change satisfies the acceptance criteria it claims (stage 1) and meets quality/convention/edge-case bars (stage 2). Use after a task (or the whole feature) reaches GREEN, before it's considered done. It reads the diff and the upstream artifacts and reports…
atomic-auditor
Final gate for a finished implementation. Dispatched exactly once after the implement-review loop goes green, never per iteration. Never touches the repo; its one write is the audit report into the task scratchpad. Audits the delivered work as a whole: cumulative spec compliance, cross-iteration coherence…