openarg_backend CLAUDE.md

openarg_backend CLAUDE.md is an instructions file for coding agents from colossus-lab/openarg_backend. It costs 3,637 tokens per session, scanned A, original, MIT.

Project instructions for an OpenArg backend, a service that collects Argentine government data and answers questions about it. They describe its software stack, hexagonal architecture, background workers, and resilience rules.

In plain words
What is it for?
Use them when modifying the API, database, data-scraping and embedding pipeline, Celery background jobs, authentication, configuration, or error handling.
Why use it?
They give coding agents the project context needed to change the backend without breaking its layers, data pipeline, or worker processing.

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/colossus-lab/openarg_backend/claude-md
Clone the repo
git clone --depth 1 https://github.com/colossus-lab/openarg_backend

Wrote 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.

agentmods badge for openarg_backend CLAUDE.md

README.md
[![agentmods](https://agentmods.dev/badge/instructions/colossus-lab/openarg_backend/claude-md.svg)](https://agentmods.dev/instructions/colossus-lab/openarg_backend/claude-md)
Your own site
<a href="https://agentmods.dev/instructions/colossus-lab/openarg_backend/claude-md"><img src="https://agentmods.dev/badge/instructions/colossus-lab/openarg_backend/claude-md.svg" alt="Measured on agentmods" height="20"></a>
Per session 3,637 This file is loaded in full into every session.
When invoked 3,637 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 $0.03637 $0.03637
Opus 5 $0.01818 $0.01818
Sonnet 5 $0.00727 $0.00727
Haiku 4.5 $0.00364 $0.00364

Measured 4d ago against content hash d679bebba94f, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

openarg_backend 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 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.

CLAUDE.md · 269 lines

How it starts

The opening of the file, as written. The whole thing — 269 lines — stays where its author put it; the contents beside it link to each section on GitHub.

OpenArg Backend

Backend service for OpenArg — AI-powered analysis of Argentine government open data. Implements a pipeline that scrapes public data portals, generates vector embeddings, caches datasets, and answers natural-language queries using LLMs.

Stack

  • Framework: FastAPI 0.115 + Uvicorn (async, UVLoop)
  • Database: PostgreSQL 16 + pgvector (HNSW indexing, 1024-dim embeddings)
  • ORM: SQLAlchemy 2.0 (async) + Alembic migrations
  • DI: Dishka 1.6 (IoC container)
  • Workers: Celery 5.4 + Redis 7 (broker + cache + results)
  • AI: AWS Bedrock Claude Haiku 4.5 (primary LLM) + Google Gemini 2.5 Flash (fallback)
  • Embeddings: AWS Bedrock Cohere Embed Multilingual v3 (1024-dim)
  • Pipeline: LangGraph (stateful graph with checkpointing)
  • HTTP: HTTPX (async client)
  • Auth: PyJWT + bcrypt
  • Rate Limiting: SlowAPI
  • Config: TOML files + Pydantic settings
  • Logging: structlog

Architecture (Hexagonal / Ports & Adapters)

src/app/
├── domain/                                      # Domain layer
│   ├── entities/                                # Dataclass entities
│   │   ├── base.py                              # BaseEntity (id, created_at, updated_at)
│   │   ├── dataset.py                           # Dataset, DatasetChunk
│   │   ├── user_query.py                        # UserQuery, AgentTask
│   │   └── query_dataset_link.py
│   ├── ports/                                   # Abstract interfaces (ABC)
│   │   ├── source/data_source.py                # IDataSource: fetch_catalog, download_dataset
│   │   ├── dataset/dataset_repository.py        # IDatasetRepository: save, get_by_id, upsert
│   │   ├── llm/llm_provider.py                  # ILLMProvider, IEmbeddingProvider
│   │   ├── search/vector_search.py              # IVectorSearch: search_datasets, index_dataset
│   │   ├── sandbox/sql_sandbox.py               # ISQLSandbox: execute_readonly
│   │   └── cache/cache_port.py                  # ICacheService: get, set, delete
│   └── exceptions/                              # Domain exceptions
│
├── infrastructure/                              # Infrastructure layer
│   ├── adapters/
│   │   ├── source/
│   │   │   ├── datos_gob_ar_adapter.py          # IDataSource → datos.gob.ar CKAN
│   │   │   └── caba_adapter.py                  # IDataSource → CABA CKAN
│   │   ├── llm/
│   │   │   ├── bedrock_llm_adapter.py           # ILLMProvider → Claude Haiku 4.5 (AWS Bedrock, primary)
│   │   │   ├── bedrock_embedding_adapter.py     # IEmbeddingProvider → Cohere Embed Multilingual v3 (1024-dim)
│   │   │   ├── gemini_adapter.py                # ILLMProvider → Gemini 2.5 Flash (Google, fallback)
│   │   │   ├── anthropic_adapter.py             # ILLMProvider → Claude Sonnet 4 (Anthropic API)
│   │   ├── search/
│   │   │   └── pgvector_search_adapter.py       # IVectorSearch → pgvector
│   │   ├── sandbox/
│   │   │   └── pg_sandbox_adapter.py            # ISQLSandbox → read-only PG queries
│   │   ├── dataset/
│   │   │   └── dataset_repository_sqla.py       # IDatasetRepository → SQLAlchemy
│   │   └── cache/
│   │       └── redis_cache_adapter.py           # ICacheService → Redis
│   ├── resilience/                              # Fault tolerance
│   │   ├── retry.py                             # @with_retry decorator (exponential backoff + jitter)
│   │   └── circuit_breaker.py                   # In-memory circuit breaker (CLOSED→OPEN→HALF_OPEN)
│   ├── monitoring/                              # Observability
│   │   ├── health.py                            # HealthCheckService (postgres, redis, ddjj, sesiones)
│   │   ├── metrics.py                           # MetricsCollector singleton (requests, connectors, cache, tokens)
│   │   └── middleware.py                        # MetricsMiddleware (ASGI)
│   ├── persistence_sqla/
│   │   ├── mappings/                            # SQLAlchemy table ↔ entity mappings
│   │   ├── alembic/versions/                    # Migration files
│   │   └── provider.py                          # DB session provider
│   └── celery/
│       ├── app.py                               # Celery app config + task routing
│       └── tasks/
│           ├── scraper_tasks.py                 # scrape_catalog, index_dataset_embedding
│           ├── collector_tasks.py               # collect_dataset (download + cache in PG)
│           ├── embedding_tasks.py               # reindex_all_embeddings
│           └── analyst_tasks.py                 # analyze_query (plan → search → gather → analyze)
│
├── presentation/http/controllers/               # API layer
│   ├── root_router.py                           # Composes all routers under /api/v1
│   ├── health/health_router.py                  # GET /health, /health/ready
│   ├── datasets/datasets_router.py              # CRUD + scrape trigger
│   ├── query/query_router.py                    # Query submission + WebSocket stream
│   ├── query/smart_query_v2_router.py           # LangGraph pipeline (POST /smart + WS /ws/smart)
│   ├── public_api/ask_router.py                 # POST /ask (public API with Bearer token auth)
│   ├── developers/developers_router.py          # API key CRUD (create/list/revoke keys)
│   ├── skills/skills_router.py                  # GET /skills (list auto-detected skills)
│   ├── sandbox/sandbox_router.py                # SQL sandbox + NL2SQL
│   ├── taxonomy/taxonomy_router.py              # Taxonomy management
│   ├── transparency/transparency_router.py      # Transparency data
│   ├── admin/tasks_router.py                    # Admin task management
│   └── monitoring/metrics_router.py             # GET /api/v1/metrics
│
└── setup/
    ├── ioc/provider_registry.py                 # Dishka providers (all DI wiring)
    ├── config/
    │   ├── settings.py                          # Pydantic settings classes
    │   └── loader.py                            # TOML config loader
    └── run.py                                   # App factory (make_app)

Read the full file on GitHub · 269 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. 4d ago First seen · 269 lines · 3,637 tokens per session scan A d679bebba94f

Subscribe to this mod's changes

openarg_backend CLAUDE.md is an instructions file published in the GitHub repository colossus-lab/openarg_backend (144 stars, last pushed yesterday), licensed MIT. It adds 3,637 tokens to every session, about $0.0182 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.

Related

Other instructions, from other repositories

vscode buildNext.instructions.md

Working notes and architecture documentation for the new esbuild-based build system in build/next. Use when making changes to the new build pipeline (transpile/bundle commands, NLS plugin, source-map handling, resource copying, or self-hosting watch tasks).

microsoft/vscode · 6,785 tokens

spec-kit AGENTS.md

AGENTS.md instructions for github/spec-kit, covering agents.md, about spec kit and specify, quickstart — add a new integration in 5 steps, integration architecture and integrationmanifest — file tracking.

github/spec-kit · 7,104 tokens

codex AGENTS.md

AGENTS.md instructions for openai/codex, covering rust/codex-rs, the codex-core crate, code review rules, crate api surface and model visible context.

openai/codex · 5,182 tokens

langchain AGENTS.md

AGENTS.md instructions for langchain-ai/langchain, covering global development guidelines for the langchain monorepo, corridor security analysis, project architecture and context, monorepo structure and development tools & commands.

langchain-ai/langchain · 4,345 tokens

vscode oss-third-party-notices.instructions.md

Instructions for microsoft/vscode, covering vs code oss third-party-notices pipeline, architecture, pipeline flow in ci, applying the notice (cutover) and fallback chain (never fail the build).

microsoft/vscode · 5,001 tokens

next.js AGENTS.md

Instructions for vercel/next.js, covering next.js development guide, codebase structure, monorepo overview, core package: packages/next and other important packages.

vercel/next.js · 7,296 tokens