Borrowing it
Nothing to install: this file belongs to FeSens/auto-arch-tournament. 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/FeSens/auto-arch-tournament/main/CLAUDE.mdgit clone --depth 1 https://github.com/FeSens/auto-arch-tournamentWrote 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/instructions/fesens/auto-arch-tournament/claude-md)<a href="https://agentmods.dev/instructions/fesens/auto-arch-tournament/claude-md"><img src="https://agentmods.dev/badge/instructions/fesens/auto-arch-tournament/claude-md/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/instructions/fesens/auto-arch-tournament/claude-md"><img src="https://agentmods.dev/badge/instructions/fesens/auto-arch-tournament/claude-md.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 | $0.02825 | $0.02825 |
| Opus 5.5 | $0.01130 | $0.01130 |
| Sonnet 5.5 | $0.00565 | $0.00565 |
| Haiku 4.5 | $0.00282 | $0.00282 |
Grade A, and why
auto-arch-tournament CLAUDE.md 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 1mo 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 — 165 lines — stays where its author put it; the contents beside it link to each section on GitHub.
CLAUDE.md — non-negotiable invariants and don't-touch list
Every item below was a real bug in the prior Chisel iteration. Each must be enforced by a check that fails loudly when violated. Do not weaken these checks to make a hypothesis pass.
Hard invariants
| # | Invariant | Enforced by |
|---|---|---|
| 1 | Top module core exposes an RVFI port set selected by cores/<target>/core.yaml's nret field (default 2). nret=1: only io_rvfi_*_0 ports exist; the single retirement always lands on channel 0. nret=2: both io_rvfi_*_0 and io_rvfi_*_1 exist; channel 0 carries the older of two simultaneous retirements, channel 1 the younger; single-retire cycles MUST place the retirement on channel 0 with io_rvfi_valid_1 = 0; all channel-1 ports MUST be driven (no X/Z, tie unused fields to '0). |
cocotb smoke test, riscv-formal wrapper (wrapper.sv for nret=2, wrapper_si.sv for nret=1) |
| 2 | rvfi_trap = 1 iff the retiring instruction is illegal per RV32IM (decoder default = illegal) |
riscv-formal ill check |
| 3 | EBREAK is the only SYSTEM (opcode 0x73) instruction the core treats as valid; ECALL / CSR / MRET trap |
dedicated decoder unit tests |
| 4 | rvfi_order strictly monotonic +1 per retirement across both channels combined (no gaps, no duplicates). When both channels retire in the same cycle, channel-0 order = N and channel-1 order = N+1. |
riscv-formal unique check |
| 5 | CPU makes forward progress under any symbolic instruction stream | riscv-formal liveness check |
| 5b | M-extension arithmetic produces RV32M-correct bit results | cocotb test_alu.py + make formal-deep |
| 6 | All memory accesses bounded to [0x00000000, 0x00100000) plus UART/markers [0x10000000, 0x10000200) |
sim emits oob:true, eval treats as failure |
| 7 | CoreMark CRCs match the canonical 2K-config perf run (TOTAL_DATA_SIZE=2000, -O3, ITERATIONS=10): seedcrc=0xe9f5, crclist=0xe714, crcmatrix=0x1fd7, crcstate=0x8e3a, crcfinal=0xfcaf |
validate_coremark_uart in fpga eval (COREMARK_EXPECTED in tools/eval/fpga.py is the source of truth) |
| 8 | CoreMark timing brackets start_time / stop_time markers — total elapsed cycles do NOT count |
MMIO writes at 0x10000100 / 0x10000104 |
| 9 | "Correct operation validated." literal must appear in CoreMark UART output | validate_coremark_uart |
| 10 | Decoder defaults isIllegal = 1; a bit is only cleared inside a validated opcode/funct match |
decoder unit tests with reserved encodings |
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.
- 1mo ago Changed · +11 lines · +212 tokens per session e09c4009156c
- 2mo ago First seen · 154 lines · 2,613 tokens per session scan A f4794cf4601d
auto-arch-tournament CLAUDE.md is an instructions file published in the GitHub repository FeSens/auto-arch-tournament (205 stars, last pushed 1mo ago), licensed Apache-2.0. It adds 2,825 tokens to every session, about $0.0113 per session on Opus 5.5. 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.
auto-arch-tournament CLAUDE.md compared with instructions from other repositories
| Mod | Type | Stars | Last push | Licence | Scan | Tokens | Works with |
|---|---|---|---|---|---|---|---|
| auto-arch-tournament CLAUDE.mdFeSens/auto-arch-tournament | Instructions file | 205 | 1mo ago | Apache-2.0 | A | 2,825 | Claude Code |
| vscode buildNext.instructions.mdmicrosoft/vscode | Instructions file | 195k | today | MIT | A | 9,719 | GitHub Copilot |
| vscode oss-third-party-notices.instructions.mdmicrosoft/vscode | Instructions file | 195k | today | MIT | A | 5,516 | GitHub Copilot |
| next.js AGENTS.mdvercel/next.js | Instructions file | 143k | today | MIT | A | 7,461 | Codex, OpenCode |
| langchain AGENTS.mdlangchain-ai/langchain | Instructions file | 149k | today | MIT | A | 4,680 | Codex, OpenCode |
| vscode copilot-instructions.mdmicrosoft/vscode | Instructions file | 195k | today | MIT | A | 3,093 | GitHub Copilot |
| vscode model-prompts.instructions.mdmicrosoft/vscode | Instructions file | 195k | today | MIT | A | 2,022 | GitHub Copilot |