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.
npx agentmods add instructions/ilanbenb/wa_llm/agents-mdgit clone --depth 1 https://github.com/ilanbenb/wa_llmWrote 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.
[](https://agentmods.dev/instructions/ilanbenb/wa_llm/agents-md)<a href="https://agentmods.dev/instructions/ilanbenb/wa_llm/agents-md"><img src="https://agentmods.dev/badge/instructions/ilanbenb/wa_llm/agents-md.svg" alt="Measured on agentmods" height="20"></a>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 | $0.00628 | $0.00628 |
| Opus 5 | $0.00314 | $0.00314 |
| Sonnet 5 | $0.00126 | $0.00126 |
| Haiku 4.5 | $0.00063 | $0.00063 |
Grade A, and why
wa_llm 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 4d 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.
How it starts
The opening of the file, as written. The whole thing — 38 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Repository Guidelines
Project Structure & Module Organization
- Source code:
src/— key packages:api/,handler/,models/,whatsapp/,utils/,config/. - Application entrypoint:
app/main.py(starts FastAPI, wiring DB, WhatsApp client). - Migrations:
migrations/withalembic.ini. - Tests are colocated with code:
src/**/test_*.py(e.g.,src/whatsapp/test_jid.py). - Extras:
notebooks/,data/.
Build, Test, and Development Commands
- Install deps:
uv sync --all-extras --active(Python 3.13). - Lint/format:
ruff checkandruff format. - Type check:
pyright. - Run tests:
pytest -qorpytest --cov=src --cov-report=xml. - Run API locally:
uv run python app/main.py(reads.envin repo root). Alt:uvicorn app.main:app --reload. - Docker stack:
docker compose up -d(services:postgres,whatsapp,web-server).
Coding Style & Naming Conventions
- Python style: 4‑space indent, f-strings, type hints for public functions, absolute imports from
src(e.g.,from handler.router import Router). - Naming: modules/functions
snake_case; classesCamelCase; constantsUPPER_CASE. - Keep functions small and async-friendly where I/O is involved.
- Enforce with
ruff(style/format) andpyright(types) before committing.
Testing Guidelines
- Frameworks:
pytest,pytest-asyncio,pytest-cov. - Test files:
test_*.pynext to modules (e.g.,src/handler/test_router.py). - Use
AsyncMock/fixtures (seesrc/test_utils/) and avoid real network/DB calls. - Coverage:
pytest --cov=src(XML written tocoverage.xml).
Commit & Pull Request Guidelines
- Use Conventional Commits when possible:
feat:,fix:,chore:, etc. Example:fix: avoid quote reply loop. - PRs must include: clear description, linked issues, validation steps, and notes on DB migrations if applicable. Attach logs/screenshots where useful.
- Gating: ensure
ruff,pyright, andpytestpass locally.
Security & Configuration Tips
- Secrets live in
.env(root). Required keys include:DB_URI,WHATSAPP_HOST,VOYAGE_API_KEY,ANTHROPIC_API_KEY,LOGFIRE_TOKEN. - DB migrations:
alembic upgrade head; create new migration withalembic revision --autogenerate -m "<msg>".
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.
- 4d ago First seen · 38 lines · 628 tokens per session scan A 59c75b66e466
wa_llm AGENTS.md is an instructions file published in the GitHub repository ilanbenb/wa_llm (166 stars, last pushed yesterday), licensed MIT. It adds 628 tokens to every session, about $0.0031 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.
Other instructions, from other repositories
go-whatsapp-web-multidevice AGENTS.md
AGENTS.md instructions for aldinokemal/go-whatsapp-web-multidevice, covering project knowledge base, structure, where to look, code map and conventions.
Cobalt CLAUDE.md
Instructions for Auties00/Cobalt, covering cobalt, build, maven modules, architecture and client hierarchy.
whatsapp-mcp AGENTS.md
Instructions for rodrigopg/whatsapp-mcp, covering whatsapp mcp server (community fork), carregar contexto antes de trabalhar, regras inegociáveis, checklist antes de abrir pr and comandos essenciais.
whatsapp-mcp AGENTS.md
AGENTS.md instructions for verygoodplugins/whatsapp-mcp, covering agents.md, repository, architecture (read first), scope rules and pr rules.
whatsapp-cli AGENTS.md
AGENTS.md instructions for vicentereig/whatsapp-cli, covering agents.md, build, run, test, code style, architecture and testing.
whatsapp-mcp CLAUDE.md
Claude Code instructions for verygoodplugins/whatsapp-mcp, a project described as: WhatsApp MCP server - Connect Claude to WhatsApp for reading and sending messages.