What is agent-system CLAUDE.md?

Project instructions for DeevsDeevs/agent-system covering coding practices, comments, Git workflow, and required use of Context7 for current library documentation. Context7 is a documentation lookup service for software libraries.

Guiding work in the specified repository, requiring documentation lookups before using external libraries, and standardizing code comments, commits, branches, and project practices.

They give agents consistent project rules and reduce the risk of using outdated library information. They also define expectations for code quality, comments, commit messages, and branch names.

agent-system CLAUDE.md — Instructions file by DeevsDeevs

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/deevsdeevs/agent-system/claude-md

Clone the repo

git clone --depth 1 https://github.com/DeevsDeevs/agent-system

Per session 560 This file is loaded in full into every session.

When invoked 560 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.

Model	Per session	Once invoked
Fable 5	$0.00560	$0.00560
Opus 5	$0.00280	$0.00280
Sonnet 5	$0.00112	$0.00112
Haiku 4.5	$0.00056	$0.00056

Measured 2d ago against content hash a0e6081f0674, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

agent-system 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 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.

CLAUDE.md · 80 lines

How it starts

The opening of the file, as written. The whole thing — 80 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Baseline Project Guidelines

MCP Tools

Context7 - MUST USE

When to use:

Always when looking up library documentation
When troubleshooting library-specific issues
Before implementing features with external dependencies

How to use:

mcp__context7__resolve-library-id - Find the library ID
mcp__context7__get-library-docs - Get documentation with the ID from step 1

Why: Ensures current, accurate documentation rather than outdated knowledge.

Code Quality

Comments

Only write comments for:

Complex algorithms that aren't immediately obvious
Non-obvious workarounds for known issues/bugs
Critical "why" explanations that code cannot express

99% of code should be self-explanatory through:

Clear, descriptive naming
Proper structure and organization
Following language best practices

Git Workflow

Commit Messages

Use Conventional Commits format:

feat: - New features
fix: - Bug fixes
refactor: - Code refactoring
docs: - Documentation changes
test: - Test updates
chore: - Maintenance tasks

Example: feat: add user authentication system

Branch Naming

Use type prefixes:

feature/ - New features
fix/ - Bug fixes
refactor/ - Code refactoring

Example: feature/user-authentication

Engineering Philosophy

Approach

Pragmatic over perfect - Ship working solutions, iterate if needed
Question assumptions - Always ask clarifying questions before building
YAGNI - Don't over-engineer; build what's needed now
KISS - Prefer simple solutions over complex ones

Communication Style

Ask questions when requirements are unclear or ambiguous
Brainstorm ideas - discuss different approaches
Find optimal approach - balance simplicity, maintainability, and effectiveness
Avoid over-engineering - question if complexity is truly necessary
Be concise - get to the point, focus on solutions

Venue Context

Use the venue-expert skill for exchange-specific context. Invoke it when:

Starting any research involving exchange data
Debugging trading system issues (feed, order, session problems)
Needing venue mechanics: order types, matching rules, session times, data formats, regulatory quirks

Read the full file on GitHub · 80 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.

2d ago First seen · 80 lines · 560 tokens per session scan A a0e6081f0674

Subscribe to this mod's changes

agent-system CLAUDE.md is an instructions file published in the GitHub repository DeevsDeevs/agent-system (40 stars, last pushed 1mo ago), licensed Apache-2.0. It adds 560 tokens to every session, about $0.0028 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 agents Git and workflow ai-agents claude-code opencode

agent-system CLAUDE.md