agentmods.dev

ai-research-arm: Instructions file for Codex

AGENTS.md

ai-research-arm AGENTS.md is an instructions file for Codex, OpenCode from guzus/ai-research-arm. It costs 1,959 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 CodexOpenCode

Written for Codex and OpenCode: reads ~/.codex or $CODEX_HOME, but also the file is AGENTS.md. Also seen: mentions CLAUDE.md; mentions Claude Code; mentions AGENTS.md.

This is guzus/ai-research-arm's own configuration. It tells 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 11repo changed 2d ago A scan 1,959 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: 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.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.svg" alt="Measured on agentmods" height="20"></a>
Per session 1,959 This file is loaded in full into every session.
When invoked 1,959 The same file — it is already loaded in full.
Security scan A 0 findings. 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 $0.01959 $0.01959
Opus 5 $0.00979 $0.00979
Sonnet 5 $0.00392 $0.00392
Haiku 4.5 $0.00196 $0.00196

Measured 2d ago against content hash aa06b3bc46ba, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-08, from the pricing page.

Security

Grade A, and why

ai-research-arm AGENTS.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 2d 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.

AGENTS.md · 142 lines

How it starts

The opening of the file, as written. The whole thing — 142 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

  • Scheduled model workflows use the Claude-harness .github/actions/agent-run directly or, for RSS, community, arXiv, Bluesky, and wiki, the trusted .github/actions/agent-dispatch. RSS/community/Bluesky reference research-editorial; arXiv/wiki reference research-editorial-secondary. Both temporarily select the registered Cursor CLI adapter (cursor-grok-4p6-fast, Grok 4.6 Fast) because the OpenCode Go monthly plan is exhausted. Host-checkout agent-run is deliberately capability-incompatible and rejected. This emergency convergence restores the three actively failing lanes but temporarily expands one Cursor failure domain across all five editorial lanes. 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"

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

    Reference: https://code.claude.com/docs/en/github-actions (action repo: https://github.com/anthropics/claude-code-action)

  • GLM-5.3 is the default fallback for ordinary Claude-harness lanes. The editorial dispatcher temporarily serves both strict routes with Cursor Grok 4.6 Fast while the OpenCode Go monthly plan is exhausted. Known credential slots are prewired for a future isolated adapter, but the current host-checkout agent-run profiles cannot be selected by this route. agent-run probes the requested provider and walks the ordered fallback.chain from data/agent-backends.json when it is unavailable (currently native Claude → Z.ai GLM). Compatibility-tested local digest synthesis/audio, primary Twitter/repair, and no-MCP AI-news lanes are explicit exceptions: their lane-local chains replace the currently unusable global secondary with isolated Cursor. Claude remains primary and currently healthy; these Cursor paths are standby fallbacks. The MCP AI-news path stays direct native Claude because Cursor denies MCP. Lanes marked "strict": true and fireworks-fallback: none runs never fall back. Set expected-paths in agent-run, or call .github/actions/require-output after deterministic commit steps, so green no-op runs do not leave the freshness watchdog stale. RSS, HN/Reddit community, arXiv, Bluesky, and wiki fail closed; only daily-digest has a deterministic publishing fallback. Check the agent/fallback step logs before drawing provider conclusions.

Read the full file on GitHub · 142 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. 2d ago Changed · +2 lines · +50 tokens per session aa06b3bc46ba
  2. 4d ago Changed · +13 lines · +152 tokens per session 2a69093ec367
  3. 8d 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 (11 stars, last pushed today), licensed MIT. It adds 1,959 tokens to every session, about $0.0098 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 2026-08-31.

aiai-agentsclaudegithub-actionsnews
Related

Other instructions, from other repositories

next.js AGENTS.md

AGENTS.md instructions for vercel/next.js, covering next.js development guide, codebase structure, monorepo overview, core package: packages/next and other important packages.

vercel/next.js · 7,296 tokens

codex AGENTS.md

AGENTS.md instructions for openai/codex, covering rust/codex-rs, the codex-core crate, code review rules, crate api surface and model visible context.

openai/codex · 5,153 tokens

vscode buildNext.instructions.md

Working notes and architecture documentation for the new esbuild-based build system in build/next. Use when making changes to the new build pipeline (transpile/bundle commands, NLS plugin, source-map handling, resource copying, or self-hosting watch tasks).

microsoft/vscode · 6,785 tokens

vscode oss-third-party-notices.instructions.md

Instructions for microsoft/vscode, covering vs code oss third-party-notices pipeline, architecture, pipeline flow in ci, applying the notice (cutover) and fallback chain (never fail the build).

microsoft/vscode · 5,001 tokens

langchain AGENTS.md

AGENTS.md instructions for langchain-ai/langchain, covering global development guidelines for the langchain monorepo, corridor security analysis, project architecture and context, monorepo structure and development tools & commands.

langchain-ai/langchain · 4,469 tokens

spec-kit AGENTS.md

AGENTS.md instructions for github/spec-kit, covering agents.md, about spec kit and specify, quickstart — add a new integration in 5 steps, integration architecture and integrationmanifest — file tracking.

github/spec-kit · 7,104 tokens