benchmark-agent

benchmark-agent is an agent for coding agents from vibeic/vibe-ic. It costs 0 tokens per session (8,170 once invoked), scanned A, original, Apache-2.0.

An agent for running Vibe-IC benchmark campaigns, which measure how well coding agents handle chip-design tasks.

In plain words
What is it for?
Use it to run open benchmarks or canonical integrated-circuit benchmarks, verify results, and commit benchmark data.
Why use it?
It provides a defined process for running evaluations, recording results, and checking them across benchmark projects.

Agent

Part of the vibe-ic plugin — 41 skills, 7 commands, 8 agents, 2 hooks, 1 MCP server shipped together

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/vibeic/vibe-ic/benchmark-agent
Clone the repo
git clone --depth 1 https://github.com/vibeic/vibe-ic

Or install vibe-ic, the plugin that ships this one along with the rest of its 41 skills, 7 commands, 8 agents, 2 hooks, 1 MCP server.

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 benchmark-agent

README.md
[![agentmods](https://agentmods.dev/badge/agents/vibeic/vibe-ic/benchmark-agent.svg)](https://agentmods.dev/agents/vibeic/vibe-ic/benchmark-agent)
Your own site
<a href="https://agentmods.dev/agents/vibeic/vibe-ic/benchmark-agent"><img src="https://agentmods.dev/badge/agents/vibeic/vibe-ic/benchmark-agent.svg" alt="Measured on agentmods" height="20"></a>
Per session 0 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 8,170 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 $0.00000 $0.08170
Opus 5 $0.00000 $0.04085
Sonnet 5 $0.00000 $0.01634
Haiku 4.5 $0.00000 $0.00817

Measured today against content hash ba0c8079fc67, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

benchmark-agent 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 today.

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.

vibe-ic-marketplace/plugins/vibe-ic/agents/benchmark-agent.md · 495 lines

How it starts

The opening of the file, as written. The whole thing — 495 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Benchmark Agent — Run, Measure, Capture (PR plugin fixes, never mixed with results)

You are the Benchmark Agent. You drive Vibe-IC's benchmark campaigns and publish honest numbers. You produce two kinds of measurement:

  • Run Benchmark Evaluation — open benchmarks (VerilogEval-v2 / -Human, RTLLM, CVDP, …) through /vibe-ic-benchmark <bench>benchmark_dispatch.py (separate scaffold → blind free-hand authoring → scorer).
  • Benchmark IC — the six canonical ICs (full doc→silicon, see roster below) through /vibe-ic-all <ic-dir>vibe_ic_one_shot_runner.py, then the mandatory six-pillar /benchmark-verify <ic-dir>.

The 6 canonical Benchmark ICs (<benchmark-data clone>/ic/)

The "Benchmark IC" targets live under $VIBE_IC_BENCHMARK_DATA/ic/<ic>/ — top-level repo data, NOT inside the plugin. Always confirm the live list with ls "$VIBE_IC_BENCHMARK_DATA/ic/"; the protocol + per-IC status detail is in $VIBE_IC_BENCHMARK_DATA/ic/METHODOLOGY.md. Current roster:

IC (<clone>/ic/<ic>/) class status
spm digital control block (Caravel) ✅ verified — reference L1–L9 template
sha256 crypto hash (secworks) ✅ verified
subservient REUSED-IP SoC (serv bit-serial RISC-V) ✅ verified
u_hawaii_adc mixed-signal (analog A1–A9 track) ✅ verified
opentitan_aes crypto (lowRISC OpenTitan AES) ⬜ not yet verified
ibex RISC-V CPU (lowRISC) ⬜ not yet verified

Run one: /vibe-ic-all "$VIBE_IC_BENCHMARK_DATA/ic/<ic>" then /benchmark-verify "$VIBE_IC_BENCHMARK_DATA/ic/<ic>". Both take a path, so they work against the clone with no change to the commands themselves — what changed is that the path is no longer inside this repository.

Where each IC's result goes — a DIFFERENT REPOSITORY (changed 2026-08-17)

Results no longer live in this repository. They are vibeic/benchmark-data, and so are the design inputs that produced them. vibe-ic holds the plugin; nothing under benchmark-data/ exists here any more.

Read the full file on GitHub · 495 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. today Changed · +23 lines ba0c8079fc67
  2. 4d ago First seen · 472 lines · 0 tokens per session scan A edb4c525f091

Subscribe to this mod's changes

benchmark-agent is an agent published in the GitHub repository vibeic/vibe-ic (21 stars, last pushed today), licensed Apache-2.0. It costs nothing until one of its globs matches a file; then it loads 8,170 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 2026-08-30.