field-agent-loop

field-agent-loop is a skill for Claude Code from vibeic/vibe-ic. It costs 211 tokens per session (10,661 once invoked), scanned A, original, Apache-2.0.

A closed-loop quality-improvement process for plugins used with integrated-circuit projects. It reruns a plugin on benchmark projects, finds gaps between source documents and generated results, and prepares a verified patch bundle.

In plain words
What is it for?
Testing plugin output against benchmark IC projects, investigating systematic problems across layers, documenting unresolved gaps, auditing fixes, and handing off verified candidate patches.
Why use it?
It turns repeated problems found during testing into a structured backlog or proposed changes, while keeping the work in a separate sandbox. A gatekeeper can then review one complete bundle instead of scattered fixes.

Skill for Claude Code

Written for Claude Code: Claude Code plugin machinery. Also seen: mentions subagents.

Needs its repository: it runs a file that does not travel with it, so clone the repository first. The line is python3 plugins/vibe-ic/programs/fix_surface_classify.py <issue|sha>.

Part of the vibe-ic plugin — 70 skills, 8 commands, 8 agents, 2 hooks, 1 MCP server shipped together

Good fit Testing plugin output against benchmark IC projects, investigating systematic problems across layers, documenting unresolved gaps, auditing fixes, and handing off verified candidate patches.

Compare 6 skills from other repositories ↓
Install

Getting it into your agent

It runs from inside its repository, so the clone comes first — what it calls does not travel with the file alone.

Clone the repo
git clone --depth 1 https://github.com/vibeic/vibe-ic
agentmods
npx agentmods add skills/vibeic/vibe-ic/field-agent-loop

Made for: Claude Code.

Or install vibe-ic, the plugin that ships this one along with the rest of its 70 skills, 8 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 field-agent-loop

README.md
[![agentmods](https://agentmods.dev/badge/skills/vibeic/vibe-ic/field-agent-loop.svg)](https://agentmods.dev/skills/vibeic/vibe-ic/field-agent-loop)
Your own site
<a href="https://agentmods.dev/skills/vibeic/vibe-ic/field-agent-loop"><img src="https://agentmods.dev/badge/skills/vibeic/vibe-ic/field-agent-loop.svg" alt="Measured on agentmods" height="20"></a>
Per session 211 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 10,661 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe. Third-party audits
  • NVIDIA SkillSpector warn 7 Sept 2026
SkillSpector: 1 finding, up to medium

These are SkillSpector’s own severities. On a checked sample its high-severity flags on skills were ~96% false positives — a documented command, a public API, a “never do X” rule — so we show them as a caution to read, not a verdict. Why →

  • medium Excessive Agency · line 185
    Skill enables autonomous high-impact decisions without human-in-the-loop verification. Critical operations (destructive commands, financial transactions, data deletion) should require explicit user confirmation.
    Fix: Add human-in-the-loop confirmation for destructive, irreversible, or high-impact operations. Never auto-execute commands that modify files, send data, or alter system state.
How audits are shown
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 5.1 $0.00211 $0.10661
Opus 5 $0.00105 $0.05331
Sonnet 5 $0.00042 $0.02132
Haiku 4.5 $0.00021 $0.01066

Measured 8d ago against content hash 4867cc7c1fe8, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-07, from the pricing page.

Security

Grade A, and why

field-agent-loop 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 8d ago.

The scan reads SKILL.md. This mod also ships 2 executable files (programs/check_closed_for_field_audit.sh, tests/test_compliance.py), listed below but not scanned — reading those needs a real analyzer, not pattern matching.

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/skills/field-agent-loop/SKILL.md · 832 lines

How it starts

The opening of the file, as written. The whole thing — 832 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Case-study notation. This skill cites the IC-A / USB-HID tester / MDV-A1101 BENCH-A reference project as concrete evidence for the rules below. The rules themselves are chip-AGNOSTIC and apply to any IC of the matching ic_class (see vibe-ic-marketplace/plugins/vibe-ic/programs/ic_class_profile.py). When you adopt this skill on a different IC, swap IC-A<your IC name> and USB-HID tester<your host-tester name>; the structural gates and rule bodies do not depend on those SKUs. See docs/design/CASE_STUDIES/IC-A_*.md for the full BENCH-A regression history.

Field-Agent Loop — Closed-Loop Plugin Quality Improvement

Purpose

The field-agent is the organic improvement engine for the vibe-ic plugin. It treats real IC benchmark projects as a test harness: re-run the plugin at HEAD, ask a fresh LLM to compare source docs against generated output, and convert every systematic gap it finds into a structured backlog issue.

Contribution-layer note. The field-agent is a Layer-1 intake role: it files a backlog (a report — or hands the gatekeeper a candidate-patch bundle) and audits landed fixes. It never edits the plugin/MCP and never pushes to main. Direct-push is the maintainer-internal (Layer-2) landing method for the gatekeeper's own fixes — not the field-agent's path. The maintainer resolves every backlog into the next plugin version.

The core-agent now self-verifies and CLOSES each issue it fixes (adding the core-closed label). The default terminal state is therefore CLOSED. The field-agent is the audit/reopen safety net: at every cron tick it re-checks the core-agent's closed issues on the actual benchmark — not on the unit test fixtures. If the fix holds on real silicon it stamps field-verified (terminal); if it does not, the field-agent reopens the issue (gh issue reopen), posts counter-evidence, and removes core-closed so the core-agent re-engages. This audit/reopen model is what kills the old wait-for-verification limbo where a fixed issue sat un-confirmed forever.

Read the full file on GitHub · 832 lines

Files

What ships with it

3 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.

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. 8d ago First seen · 832 lines · 211 tokens per session scan A 4867cc7c1fe8

Subscribe to this mod's changes

field-agent-loop is a skill published in the GitHub repository vibeic/vibe-ic (23 stars, last pushed today), licensed Apache-2.0. It adds 211 tokens to every session and 10,661 once invoked, about $0.0011 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 2026-08-30.

Related

Other skills, from other repositories

analog-verify

Pre-simulation review and Spectre simulation verification for analog circuits. Reviews circuit netlist and testbench, runs simulation, produces margin report. Use after analog-design completes a netlist.

Arcadia-1/analog-agents · 41 tokens

step6-verify

Step 6 · Per-module simulation verification (amnesia audit → human-confirmed plan → autonomous local-fix loop → full re-verify; escalate only when local options are exhausted).

raisoninme/boardless-pcb · 42 tokens

boardrepo

Read and review real PCB projects on BoardRepo. Search published hardware designs, open a board's schematic connectivity, bill of materials and files, run KiCad's DRC and ERC, and check a design against a fabrication house's limits. Use whenever the user names a BoardRepo board or URL, asks to find a published board…

flintt-dev/boardrepo-plugin · 156 tokens

analog-netlist-crawl

Crawl and analyze post-layout parasitic netlists without running SPICE. Answers "what's the effective resistance from node A to node B across this massive R mesh?", "inside the VREFN mesh, which device pins are electrically farthest apart?", "which nets have the worst coupling?", "where does settling bottleneck?" — by…

Arcadia-1/analog-agents · 291 tokens

analog-design

Transistor-level circuit design for one analog sub-block. Produces Spectre netlist with hand-calculation rationale. Use when designing a specific circuit block after architecture is defined.

Arcadia-1/analog-agents · 38 tokens

analog-pipeline

MANDATORY — MUST load this skill when the user mentions: OTA, ADC, PLL, comparator, bandgap, LDO, amplifier, opamp, or any analog/mixed-signal IC design task. Full analog design pipeline: spec -> architecture -> design -> verify -> deliver. Orchestrates analog-decompose, analog-behavioral, analog-design…

Arcadia-1/analog-agents · 94 tokens