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/bthink-bgu/bpjs/claude-mdgit clone --depth 1 https://github.com/bThink-BGU/BPjsWrote 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/bthink-bgu/bpjs/claude-md)<a href="https://agentmods.dev/instructions/bthink-bgu/bpjs/claude-md"><img src="https://agentmods.dev/badge/instructions/bthink-bgu/bpjs/claude-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 5.1 | $0.01144 | $0.01144 |
| Opus 5 | $0.00572 | $0.00572 |
| Sonnet 5 | $0.00229 | $0.00229 |
| Haiku 4.5 | $0.00114 | $0.00114 |
Grade A, and why
BPjs 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 6d 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 — 90 lines — stays where its author put it; the contents beside it link to each section on GitHub.
CLAUDE.md
This file provides guidance to Claude Code (claude.ai/code) when working with code in this repository.
What is BPjs
BPjs is a Java framework for executing and verifying Behavioral Programs (BP) — concurrent JavaScript programs (b-threads) that synchronize at bp.sync() points via events. The core model: each b-thread declares what events it requests, waits for, blocks, and what should interrupt it. An event selection strategy resolves these declarations to choose which event fires next.
The framework supports two operating modes: execution (run one trace) and verification (DFS over all possible traces to find violations).
Build & Test Commands
# Build
mvn clean package
# Run all tests
mvn test
# Run a specific test class
mvn test -Dtest=GeneralTest
# Run a specific test method
mvn test -Dtest=GeneralTest#someMethodName
# Build standalone uber-jar
mvn clean package -P uber-jar
# Generate coverage report (output: target/coverage-reports/index.html)
mvn clean test jacoco:report
Test framework: JUnit 4 with Hamcrest. Tests live under src/test/java and corresponding JS fixtures under src/test/resources.
Architecture
Core Loop: Execution
BProgramRunner drives the execution cycle:
- All registered b-threads reach
bp.sync()— each submits aSyncStatement EventSelectionStrategypicks one selectable event (requested but not blocked)- The chosen event is fired: waiting b-threads are resumed, interrupted b-threads are killed
- Repeat until no b-threads remain or no event can be selected
Core Loop: Analysis/Veification
DfsBProgramVerifier explores all program traces by cloning BProgramSyncSnapshot at each state and performing a DFS over he BProgram's state space.
VisitedStateStore implementations (snapshot-level or b-thread-level) detect visited nodes on that state space, and prevent the DFS from getting top an infinite loops when traversing cycles.
ExecutionTraceInspection objects define what constitutes a violation (e.g., HOT_SYSTEM, HOT_BTHREADS, assertion failures).
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.
- 6d ago First seen · 90 lines · 1,144 tokens per session scan A 52d1220ad24c
BPjs CLAUDE.md is an instructions file published in the GitHub repository bThink-BGU/BPjs (45 stars, last pushed 4mo ago), licensed MIT. It adds 1,144 tokens to every session, about $0.0057 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
fonline AGENTS.md
AGENTS.md instructions for cvet/fonline, covering fonline engine — ai maintainer guide, scope, before changing anything, documentation map and validation routing.
advjs AGENTS.md
Instructions for YunYouJun/advjs, covering agents.md, project overview, common commands, run a single unit test file and dev specific packages.
advjs CLAUDE.md
Instructions for YunYouJun/advjs, covering claude.md, project overview, common commands, run a single unit test file and dev specific packages.
Kokoro-Engine CLAUDE.md
Claude Code instructions for chyinan/Kokoro-Engine, covering claude.md, git 规范, 项目概述, 常用命令 and 安装依赖.
FlingEngine AGENTS.md
Instructions for flingengine/FlingEngine, covering agents.md, what this project is, build, tests and coding style.
FlingEngine CLAUDE.md
Instructions for flingengine/FlingEngine: See @AGENTS.md for project guidance (build, tests, coding style, architecture).