bandiradar: Instructions file for Claude Code

CLAUDE.md

bandiradar CLAUDE.md is an instructions file for Claude Code from mayai-it/bandiradar. It costs 7,415 tokens per session, scanned A, original, MIT.

A set of project instructions for bandiradar, an open-source engine that monitors Italian public funding opportunities such as grants, tenders, and incentives. It explains the repository's purpose, boundaries, architecture, and information sources.

In plain words
What is it for?
It is for guiding agents when they work on funding-opportunity models, data sources, matching companies to opportunities, the command-line interface, and the MCP server.
Why use it?
It prevents coding agents from confusing the open-source engine with the separate private product or adding features that do not belong in this repository.

Instructions file for Claude Code

Written for Claude Code: the file is CLAUDE.md. Also seen: mentions CLAUDE.md.

This is mayai-it/bandiradar's own configuration. It tells Claude Code how to work on bandiradar itself, so it is not a mod to install elsewhere. Copy it as a starting point and replace the rules that are about this project. Everything bandiradar configures →

Reuse

Borrowing it

Nothing to install: this file belongs to mayai-it/bandiradar. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.

Copy the file
curl -O https://raw.githubusercontent.com/mayai-it/bandiradar/main/CLAUDE.md
Clone the repo
git clone --depth 1 https://github.com/mayai-it/bandiradar

Made for: Claude Code.

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 bandiradar CLAUDE.md

README.md
[![agentmods](https://agentmods.dev/badge/instructions/mayai-it/bandiradar/claude-md/github.svg)](https://agentmods.dev/instructions/mayai-it/bandiradar/claude-md)
Your own site
<a href="https://agentmods.dev/instructions/mayai-it/bandiradar/claude-md"><img src="https://agentmods.dev/badge/instructions/mayai-it/bandiradar/claude-md/github.svg" alt="Measured on agentmods" height="20"></a>

Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.

agentmods 80×15 button for bandiradar CLAUDE.md

Your own site · 80×15
<a href="https://agentmods.dev/instructions/mayai-it/bandiradar/claude-md"><img src="https://agentmods.dev/badge/instructions/mayai-it/bandiradar/claude-md.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 7,415 This file is loaded in full into every session.
When invoked 7,415 The same file — it is already loaded in full.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.
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 5.1 $0.07415 $0.07415
Opus 5 $0.03707 $0.03707
Sonnet 5 $0.01483 $0.01483
Haiku 4.5 $0.00741 $0.00741

Measured 8d ago against content hash 33a771b31bd6, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-08, from the pricing page.

Security

Grade A, and why

bandiradar 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 8d 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 · 403 lines

How it starts

The opening of the file, as written. The whole thing — 403 lines — stays where its author put it; the contents beside it link to each section on GitHub.

CLAUDE.md — bandiradar (open core)

Internal status, roadmap and strategy live in ROADMAP.local.md (gitignored, local only — read it first if present).

Project memory for coding agents. Read this fully before editing. The deep design lives in ARCHITECTURE.md; this file is the operational contract.

One-liner

Open-source engine that monitors Italian public funding opportunities (tenders, grants, incentives), normalizes them into one canonical Opportunity model, and ranks them against a company Profile with a two-stage matcher.

What this repo IS / IS NOT

  • IS (open, MIT): engine, Source framework, reference adapters (ANAC, later incentivi.gov.it), two-stage matcher, CLI, MCP server.
  • IS NOT (lives in private bandiradar-pro): web dashboard, premium/regional adapters, delivery channels (WhatsApp/email/alerts), multi-tenant, hosting.
  • Boundary rule: single-user/local = here. Managed/multi-client/delivery = pro.
  • Never add dashboard, paid, or delivery-channel code to this repo.

Architecture map (module → responsibility)

src/bandiradar/
  models.py        # pydantic: Opportunity, RawDoc, Profile, Match. THE contract.
  sources/
    base.py        # Source Protocol + registry
    anac.py        # ANAC/PNCP (OCDS) historical adapter + fixture
    anac_pvl.py    # ANAC PVL — OPEN tenders, live, no-creds + fixture
    heal.py        # LLM crawl-recipe healer (re-derives a drifted recipe)
    llm_scraper.py # reusable LLM HTML extractor; re-exports the crawl spine
    wordpress.py   # reusable WP-REST base (lazio, sicilia configs)
    plone.py       # reusable Plone `Bando` base (emilia_romagna config)
    sicilia.py     # EuroInfoSicilia FESR/FSC (WP base + categories filter)
    emilia_romagna.py # ER Politiche territoriali (Plone Bando)
    trentino.py    # Prov. Trento FEASR — CKAN open-data CSV adapter
    veneto.py      # Veneto SIU — LlmScraperSource (landing, HTML regex-recipe heal)
    piemonte.py    # Piemonte Drupal — LlmScraperSource (Views, HTML regex-recipe heal)
    puglia.py      # Puglia PR 21-27 — LlmScraperSource (Liferay fragment, badge; detect-only)
    sardegna.py    # Sardegna Impresa — LlmScraperSource (Views, HTML regex-recipe heal)
    fvg.py         # FVG bandi module — LlmScraperSource (contributi filter in regex; recipe heal; relay in CI)
    campania.py    # Sviluppo Campania — LlmScraperSource (widget anchors, title_template; recipe heal)
    calabria.py    # Calabria Europa — LlmScraperSource (WP-REST CPT, JSON recipe heal)
    basilicata.py  # Portalebandi — LlmScraperSource (WP-REST CPT, JSON recipe heal)
    liguria.py     # Liguria publiccompetition — LlmScraperSource (POST+CSRF fetch, regex recipe heal)
  cpv.py           # CPV Italian-label → 8-digit code resolver (pure, offline)
  trust.py         # trust spine: deterministic validation of LLM extractions (pure)
  crawl.py         # self-healing crawl spine (stdlib: recipes + drift + golden)
  recipe_store.py  # per-source CrawlRecipe overrides + golden (CONFIG, not code)
  matching/
    prefilter.py   # Stage 1: pure deterministic filter
    relevance.py   # Stage 2: LLM scorer (+ offline fallback + cache)
    llm.py         # provider-agnostic LLM client
    prompts.py     # prompt templates
  storage.py       # SQLite store, dedupe, change detection
                   #   + crawl_recipes + crawl_golden tables (self-healing crawl)
  core.py          # service layer that orchestrates the pipeline
  evaluation.py    # matching-quality eval: pure metrics + run over eval corpus
  cli.py           # Typer CLI (thin)
  mcp_server.py    # FastMCP server (thin)
  resources.py     # importlib.resources access to packaged data (below)
  data/            # PACKAGED runtime data (ships in the wheel)
    fixtures/      # recorded source payloads for offline tests / --sample
    profiles/      # example profiles (mayai.yaml, manifattura.yaml, …)
    eval/          # labelled eval set: opportunities.jsonl + gold.yaml
    cpv_it.json    # official EU CPV 2008 vocabulary (Italian label → code)
    comuni_it.json # ISTAT comuni table (region resolution for anac_pvl)
tests/
scripts/
  monitor_status.py    # live-monitor STATUS.md generator (pure, offline, tested)
.github/workflows/
  monitor.yml          # daily self-maintaining monitor (keyless; key→LLM+healer)

Runtime data lives INSIDE the package and is reached via bandiradar.resources (importlib.resources) — never Path(__file__).parents[...] — so --sample and the bundled example profiles work from a pip-installed wheel, not only a checkout. Interfaces (cli.py, mcp_server.py) are THIN — no business logic. All logic lives in core.py, sources/, matching/, storage.py.

Read the full file on GitHub · 403 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. 8d ago First seen · 403 lines · 7,415 tokens per session scan A 33a771b31bd6

Subscribe to this mod's changes

bandiradar CLAUDE.md is an instructions file published in the GitHub repository mayai-it/bandiradar (2 stars, last pushed 2mo ago), licensed MIT. It adds 7,415 tokens to every session, about $0.0371 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 2026-08-31.

Related

Other instructions, from other repositories

next.js AGENTS.md

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

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 · 5,153 tokens

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

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

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 · 4,469 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