agentmods.dev

ai-research-arm: Instructions file for Claude Code

AGENTS.md

ai-research-arm AGENTS.md is an instructions file for Claude Code, Codex, OpenCode from guzus/ai-research-arm. It costs 2,365 tokens per session, scanned A, original, MIT.

Agent instructions for AI Research Arm, a system that runs scheduled research workflows from sources such as RSS feeds, arXiv and community sites. They direct agents to the full operating guide and distinguish runnable features from boundaries.

In plain words
What is it for?
Finding the correct project documentation, understanding scheduled research jobs and selecting the permitted execution path for each supported source.
Why use it?
They help an agent choose the approved workflow and research adapter instead of using an incompatible or unregistered route.

Instructions file for Claude CodeCodexOpenCode

Written for Claude Code and Codex and OpenCode: PreToolUse hook event, but also the file is AGENTS.md. Also seen: mentions CLAUDE.md; mentions Claude Code; reads ~/.codex paths.

This is guzus/ai-research-arm's own configuration. It tells Claude Code, Codex and OpenCode how to work on ai-research-arm itself, so it is not a mod to install elsewhere. Copy it as a starting point and replace the rules that are about this project. Everything ai-research-arm configures →

not rated 14repo +1 changed yesterday A scan 2,365 tokens original MIT
View source ↗ guzus/ai-research-arm ara.guzus.xyz
Reuse

Borrowing it

Nothing to install: this file belongs to guzus/ai-research-arm. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.

Copy the file
curl -O https://raw.githubusercontent.com/guzus/ai-research-arm/main/AGENTS.md
Clone the repo
git clone --depth 1 https://github.com/guzus/ai-research-arm

Made for: Claude Code, Codex, OpenCode.

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 ai-research-arm AGENTS.md

README.md
[![agentmods](https://agentmods.dev/badge/instructions/guzus/ai-research-arm/agents-md/github.svg)](https://agentmods.dev/instructions/guzus/ai-research-arm/agents-md)
Your own site
<a href="https://agentmods.dev/instructions/guzus/ai-research-arm/agents-md"><img src="https://agentmods.dev/badge/instructions/guzus/ai-research-arm/agents-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.

agentmods 80×15 button for ai-research-arm AGENTS.md

Your own site · 80×15
<a href="https://agentmods.dev/instructions/guzus/ai-research-arm/agents-md"><img src="https://agentmods.dev/badge/instructions/guzus/ai-research-arm/agents-md.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 2,365 This file is loaded in full into every session.
When invoked 2,365 The same file — it is already loaded in full.
Security scan A 1 finding. A grade says what 26 rules found in the file — not that it is safe.
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 $0.02365 $0.02365
Opus 5.5 $0.00946 $0.00946
Sonnet 5.5 $0.00473 $0.00473
Haiku 4.5 $0.00236 $0.00236

Measured yesterday against content hash cbbf991b1bea, method: parsed. Prices are Anthropic first-party input rates as of 2026-10-07, from the pricing page.

Security

Grade A, and why

ai-research-arm AGENTS.md scanned grade A with 1 finding 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.

Makes network callslowCapability

Not a fault in itself. Listed so you know the mod talks to something, and to what.

add/commit, one fixed curl prefix, named uv scripts), and a PreToolUse
AGENTS.md · 165 lines

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.

AGENTS.md

Agents working in this repo: read CLAUDE.md — it is the full operating guide (pipeline architecture, workflows, output locations, authentication, and the load-bearing rules). This file is a short pointer plus the few genuinely agent-specific notes.

New to the repo and not operating the live pipeline? Start with README.md — its Quickstart runs with no accounts, and it documents the License and the runnable-vs-not boundary.

Agent-specific notes

  • Langfuse research pilot and RSI: follow docs/research-observability.md. Capture attempts before transcript overwrite, keep Langfuse keys in separate reporting jobs, and treat missing telemetry as unknown. Retain RSI candidates only after frozen matched evaluation and reserved-case transfer; trace installation alone does not demonstrate improvement.

  • Scheduled model workflows use the Claude-harness .github/actions/agent-run directly or, for RSS, community, arXiv, Bluesky, wiki, Korean translation, and Aradio, the trusted .github/actions/agent-dispatch. RSS/community/Bluesky reference research-editorial; arXiv/wiki reference research-editorial-secondary. Since 2026-11-05 all strict routes use Cursor Grok 4.6 Fast after canary 41962718497 passed. Claude executions hit a weekly-limit HTTP 429 before doing work; OpenCode canary 41962723359 failed before its GLM probe, so OpenCode recovery remains unverified. All strict routes temporarily share Cursor. Restore research-editorial to opencode-glm-5p3-flash only after its registered editorial canary passes. Host-checkout agent-run remains capability-incompatible. Direct Claude workflows may still use anthropics/claude-code-action@v1; when they do, pass the model via claude_args, never as a separate model: input:

    # Correct (v1)
claude_args: "--model claude-sonnet-5-5"

# Wrong (do not use a separate model input here)
model: <versioned-model-name>

Read the full file on GitHub · 165 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 Changed · +23 tokens per session cbbf991b1bea
  2. 5d ago Changed · +9 lines · +205 tokens per session b383e5430790
  3. 12d ago Changed · +14 lines · +176 tokens per session 2e71547d04a2
  4. 1mo ago Changed · +2 tokens per session bfd73444c41e
  5. 2mo ago Changed · +2 lines · +50 tokens per session aa06b3bc46ba
  6. 2mo ago Changed · +13 lines · +152 tokens per session 2a69093ec367
  7. 2mo ago First seen · 127 lines · 1,757 tokens per session scan A 839c9edaf05b

Subscribe to this mod's changes

ai-research-arm AGENTS.md is an instructions file published in the GitHub repository guzus/ai-research-arm (14 stars, last pushed today), licensed MIT. It adds 2,365 tokens to every session, about $0.0095 per session on Opus 5.5. A static security scan graded it A with 1 finding (makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.

aiai-agentsclaudegithub-actionsnews
Related

ai-research-arm AGENTS.md compared with instructions from other repositories

ai-research-arm AGENTS.md next to 6 instructions from other repositories
ModTypeStarsLast pushLicenceScanTokensWorks with
ai-research-arm AGENTS.mdguzus/ai-research-armInstructions file14todayMITA2,365Claude Code, Codex, OpenCode
vscode buildNext.instructions.mdmicrosoft/vscodeInstructions file195kyesterdayMITA9,719GitHub Copilot
next.js AGENTS.mdvercel/next.jsInstructions file143kyesterdayMITA7,461Codex, OpenCode
langchain AGENTS.mdlangchain-ai/langchainInstructions file149kyesterdayMITA4,680Codex, OpenCode
vscode oss-third-party-notices.instructions.mdmicrosoft/vscodeInstructions file195kyesterdayMITA5,516GitHub Copilot
vscode copilot-instructions.mdmicrosoft/vscodeInstructions file195kyesterdayMITA3,093GitHub Copilot
deepseek-harness AGENTS.mddeepseek-ai/deepseek-harnessInstructions file274k2d agoMITA4,313Codex, OpenCode