LAP CLAUDE.md

A guide for LAP, a tool that converts API descriptions into a shorter format for AI agents. API descriptions define how software services can be called, using formats such as OpenAPI, GraphQL, AsyncAPI, Protobuf, or Postman.

In plain words
What is it for?
Use it to compile API specifications, convert them back, compare API changes, parse LAP files, run tests, or use its Python and TypeScript integrations.
Why use it?
Shorter API descriptions give AI agents less text to process while retaining the described API information.

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/lap-platform/lap/claude-md
Clone the repo
git clone --depth 1 https://github.com/Lap-Platform/LAP
Per session 971 This file is loaded in full into every session.
When invoked 971 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 $0.00971 $0.00971
Opus 5 $0.00485 $0.00485
Sonnet 5 $0.00194 $0.00194
Haiku 4.5 $0.00097 $0.00097

Measured 3d ago against content hash 9c2448b9440e, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

LAP 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 · 86 lines

How it starts

The opening of the file, as written. The whole thing — 86 lines — stays where its author put it; the contents beside it link to each section on GitHub.

LAP - Lean API Platform

What This Is

LAP compiles API specs (OpenAPI, GraphQL, AsyncAPI, Protobuf, Postman) into LAP -- a token-efficient format for AI agents. Median 2.8x compression across 162 real-world specs.

Quick Reference

# Install
pip install -e ".[dev]"

# Run tests
python -m pytest tests/ -q

# CLI entry point
lapsh compile examples/verbose/openapi/petstore.yaml -o petstore.lap

Project Structure

  • lap/ -- Top-level Python package (installed as lapsh)
    • lap/core/compilers/ -- Format-specific compilers (openapi, graphql, asyncapi, protobuf, postman, lap)
    • lap/core/formats/ -- Data models (LAP v0.2, LAP v0.1)
    • lap/core/parser.py -- LAP text to Python objects (proves losslessness)
    • lap/core/converter.py -- LAP to OpenAPI roundtrip
    • lap/core/differ.py -- Semantic API diff engine
    • lap/core/utils.py -- Shared utilities (token counting, file reading)
    • lap/cli/main.py -- CLI with 15+ subcommands
  • integrations/ -- LangChain, CrewAI, OpenAI, MCP bridges
  • sdks/python/ -- Python SDK (thin wrapper around lap.core)
  • sdks/typescript/ -- TypeScript/npm SDK (@lap-platform/lapsh)
  • tests/ -- pytest suite (11 test files)
  • benchmarks/ -- Skill token benchmarks
  • assets/ -- Chart PNGs for docs
  • examples/verbose/ -- 162 real-world API specs (36MB corpus), organized by format
  • examples/lap/ -- Pre-compiled LAP output, organized by format

Architecture

API Spec (YAML/JSON/SDL/proto)
  -> Format compiler (lap/core/compilers/*.py)
  -> LAP data model (lap/core/formats/*.py)
  -> .to_lap(lean=True/False) text output
  -> Parser (lap/core/parser.py) for roundtrip validation
  -> Converter/Differ for analysis

Tech Stack

  • Python 3.10+
  • Dependencies: pyyaml, tiktoken, rich (dev)
  • Tests: pytest
  • Package: setuptools (pyproject.toml)
  • License: Apache 2.0

Key Conventions

  • All imports use the lap package namespace (e.g., from lap.core.formats.lap import LAPSpec)
  • Framework integrations handle ImportError gracefully -- no hard deps on LangChain/CrewAI/etc.
  • Token counting uses tiktoken with gpt-4o model, falls back to len(text)//4
  • CLI entry point: lap.cli.main:main
  • Tests run from project root: python -m pytest tests/ -q
  • All compilers follow the same pattern: take spec input, return format-specific data model, call .to_lap()
  • All file reads use encoding='utf-8' (Windows cp1255 fix)

Read the full file on GitHub · 86 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. 3d ago First seen · 86 lines · 971 tokens per session scan A 9c2448b9440e

Subscribe to this mod's changes

LAP CLAUDE.md is an instructions file published in the GitHub repository Lap-Platform/LAP (346 stars, last pushed 16d ago), licensed Apache-2.0. It adds 971 tokens to every session, about $0.0049 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.