agentmods.dev

agent-collab CLAUDE.md

Project-specific instructions for agent-collab, a tool that runs Codex and Cursor as background coding workers inside Claude Code.

In plain words
What is it for?
Use it when delegating code analysis, architecture checks, security reviews, refactors, implementation work, or pre-merge reviews in that project.
Why use it?
It gives agents shared guidance about which worker and model to use for different tasks, including deeper analysis and faster implementation.

Instructions file

32 18d ago A scan 328 tokens original Apache-2.0
View the source file zebbern/agent-collab
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 instructions/zebbern/agent-collab/claude-md
Clone the repo
git clone --depth 1 https://github.com/zebbern/agent-collab
Per session 328 This file is loaded in full into every session.
When invoked 328 The same file — it is already loaded in full.
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 $0.00328 $0.00328
Opus 5 $0.00164 $0.00164
Sonnet 5 $0.00066 $0.00066
Haiku 4.5 $0.00033 $0.00033

Measured 3d ago against content hash 9c8c179b21aa, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

agent-collab 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 3d 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.

CLAUDE.md · 24 lines

What it actually says

CLAUDE.md

@AGENTS.md

Claude-specific notes on top of the shared agent guide:

  • Model delegation (repo-specific working preferences, not plugin defaults):
    • Deep code analysis, architecture validation, security review, and high-stakes refactors → Codex task/rescue with --profile deep (resolves to --model gpt-5.6-sol --effort xhigh); escalate to --profile deep --effort max when a verdict still feels shallow.
    • Fast iteration, everyday implementation, scaffolding, and the standing pre-merge reviews → Cursor with --model cursor-grok-4.5-high-fast.
    • Escalation rule: when a task that started on the Cursor model turns out to need deep analysis or security judgement, hand it to Codex --profile deep rather than pushing the faster model further.
    • Both model ids and the max effort tier were live-verified against the real CLIs on 2026-08-05 (codex-cli 0.146.0); re-verify before assuming newer names. The deep/fast profile table (plugins/codex/scripts/ codex-companion.mjs) is grounded in these same ids.
  • Key references (read on demand, not imported): README.md for the user-facing docs, plugins/*/CHANGELOG.md for history, tests/chassis-drift.test.mjs for the current divergence pins.
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. 3d ago First seen · 24 lines · 328 tokens per session scan A 9c8c179b21aa

Subscribe to this mod's changes

agent-collab CLAUDE.md is an instructions file published in the GitHub repository zebbern/agent-collab (32 stars, last pushed 18d ago), licensed Apache-2.0. It adds 328 tokens to every session, about $0.0016 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.

Building agentsagents-teamaiai-agentsai-agents-cliai-agents-frameworkai-tools
Related

Other instructions, from other repositories

agentflow-cli CLAUDE.md

Instructions for 10xHub/agentflow-cli, covering agentflow-api (api server + cli) — engineering guide, what this package is, package layout, cli commands (verified against cli/main.py) and agentflow.json (the config contract).

10xHub/agentflow-cli · 3,664 tokens

AutoAgents AGENTS.md

Instructions for liquidos-ai/AutoAgents, covering autoagents repository guidelines, project structure & module organization, code quality, naming conventions and error handling.

liquidos-ai/AutoAgents · 880 tokens

ai-workflow AGENTS.md

Instructions for cunhaax/ai-workflow, covering ai workflow template, rules — non-negotiable, project overview, commands and architecture.

cunhaax/ai-workflow · 2,602 tokens

trackly-cli CLAUDE.md

Claude Code instructions for trackly-app/trackly-cli, covering trackly-cli, tech stack, backend production source of truth, directory structure and key commands.

trackly-app/trackly-cli · 3,195 tokens

McpServer AGENTS.md

AGENTS.md instructions for sharpninja/McpServer, covering agent instructions, session start, rules, byrd test gate and where things live.

sharpninja/McpServer · 1,993 tokens

genesis-architect CLAUDE.md

Claude Code instructions for maioio/genesis-architect, covering genesis architect - claude code instructions, project structure, how this skill works, development workflow and key constraints (enforced in skill.md).

maioio/genesis-architect · 582 tokens