braid CLAUDE.md

Development instructions for Braid, a toolkit for creating custom LangGraph agents. LangGraph is a framework for building applications whose steps and decisions are represented as a graph.

In plain words
What is it for?
Use them when developing Braid agents, especially agents that call external services through the project's built-in integrations. They also define trigger phrases for planning and production-enhancement phases.
Why use it?
They give the coding agent a repeatable process for reviewing documentation and choosing direct integrations when creating agents for the project.

Instructions file

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/braid-ink/braid/claude-md
Clone the repo
git clone --depth 1 https://github.com/braid-ink/braid
Per session 1,794 This file is loaded in full into every session.
When invoked 1,794 The same file β€” it is already loaded in full.
Security scan B 2 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 $0.01794 $0.01794
Opus 5 $0.00897 $0.00897
Sonnet 5 $0.00359 $0.00359
Haiku 4.5 $0.00179 $0.00179

Measured 2d ago against content hash ffe033a009ff, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade B, and why

braid CLAUDE.md scanned grade B with 2 findings against 26 rules in 11 categories β€” prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency β€” 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.

Sends data to an external URLmediumData exfiltration

A POST to an outside endpoint may be telemetry or may be exfiltration; either way the mod talks to somewhere, and you should know where.

response = requests.post( "https://api.newservice.com/action",

Makes network callslowCapability

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

response = requests.post(
CLAUDE.md Β· 218 lines

How it starts

The opening of the file, as written. The whole thing β€” 218 lines β€” stays where its author put it; the contents beside it link to each section on GitHub.

Claude Assistant Guide for Braid Development

This file provides essential guidance for Claude when working on Braid projects, focusing on creating custom LangGraph agents using our built-in toolkit and integrations.

🎯 STRUCTURED AGENT DEVELOPMENT PROCESS

When users request agent development, follow the Agent Development Roadmap for consistent, high-quality results.

Trigger Commands

  • "Please prepare to create a LangGraph agent by reviewing the docs and examples" β†’ Execute Phase 1: Discovery & Planning
  • "Optional Pro Pack and production deployment" β†’ Execute Phase 2: Production Enhancement

These commands ensure systematic review of all critical documentation, proper template selection, and adherence to production best practices.

πŸš€ Creating New Agents with Direct API Integrations

Primary Approach: Use direct API integrations from core/integrations/ for all new agents. This provides simple, reliable access to external services.

When a user requests a new agent, follow this recommended approach:

1. Use Direct API Integrations

Primary Approach: Use direct API integrations from core/integrations/ for all new agents.

# Available direct integrations
core/integrations/
β”œβ”€β”€ xero/tools.py         # Direct Xero API integration
β”œβ”€β”€ notion/tools.py       # Direct Notion API integration  
β”œβ”€β”€ perplexity/tools.py   # Direct Perplexity API integration
β”œβ”€β”€ gworkspace/tools.py   # Google Workspace integration
└── slack/tools.py        # Slack integration

2. Standard Direct Integration Process

  1. Use Existing Integrations: Import tools from core/integrations/
  2. Configure Environment: Set API keys in .env file
  3. Import and Use: Import tools directly in your agent code
  4. Test Integration: Test API calls with direct tool invocation
  5. Handle Errors: Add graceful fallbacks for API failures

3. Available Direct Integrations

Integration Tools Use Cases
Xero get_xero_profit_and_loss, get_xero_balance_sheet, get_xero_trial_balance Financial data, accounting reports
Notion create_notion_page, get_notion_page, update_notion_page Documentation, report generation
Perplexity perplexity_search, perplexity_market_research, perplexity_ask Real-time web research, market insights
Google Workspace Gmail, Calendar, Sheets tools Email, scheduling, data management
Slack Message, channel, user tools Team communication, notifications

Read the full file on GitHub Β· 218 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 First seen Β· 218 lines Β· 1,794 tokens per session scan B ffe033a009ff

Subscribe to this mod's changes

braid CLAUDE.md is an instructions file published in the GitHub repository braid-ink/braid (11 stars, last pushed 1y ago), licensed Apache-2.0. It adds 1,794 tokens to every session, about $0.0090 per session on Opus 5. A static security scan graded it B with 2 findings (sends data to an external url, makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.