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/tornlux/unrealbridge/claude-mdgit clone --depth 1 https://github.com/TornLux/UnrealBridgeWrote 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/tornlux/unrealbridge/claude-md)<a href="https://agentmods.dev/instructions/tornlux/unrealbridge/claude-md"><img src="https://agentmods.dev/badge/instructions/tornlux/unrealbridge/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 | $0.03525 | $0.03525 |
| Opus 5 | $0.01762 | $0.01762 |
| Sonnet 5 | $0.00705 | $0.00705 |
| Haiku 4.5 | $0.00352 | $0.00352 |
Grade A, and why
UnrealBridge 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 — 145 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.
Project Overview
UnrealBridge is a TCP bridge between external tools (Claude Code) and Unreal Engine 5.3+ (verified clean BuildPlugin against 5.3.2 / 5.4.4 / 5.5.4 / 5.6.1 / 5.7.1 / 5.8.0 via tools/build_matrix.py; some libraries are gated to 5.7+ and a few inline shims handle 5.3 and 5.8 — see docs/version-compatibility.md). UE 5.2 and earlier are not supported. It consists of:
- A UE Editor plugin (
Plugin/UnrealBridge/) that runs a TCP server inside the editor - A Python CLI client (
.claude/skills/unreal-bridge/scripts/bridge.py) used by theunreal-bridgeskill - API reference docs and helper scripts for querying/manipulating UE assets via Python
Key Commands
Sync plugin and matching skills to a UE project:
sync_project.bat D:\Path\To\YourProject
Mirrors Plugin/UnrealBridge/ plus .claude/skills/unreal-bridge/ into the target project's plugin, .agents, and .claude destinations. Local plugin build outputs are retained. sync_plugin.bat is a legacy command-name alias with the same project-root argument.
Test bridge connection:
python .claude/skills/unreal-bridge/scripts/bridge.py ping
Execute Python in UE:
python .claude/skills/unreal-bridge/scripts/bridge.py exec "print('hello')"
python .claude/skills/unreal-bridge/scripts/bridge.py exec-file script.py
Architecture
TCP Protocol
Length-prefixed JSON over TCP on an OS-assigned port (default bind 127.0.0.1, port auto-allocated at startup). Clients find the editor via the UDP discovery service (host-local by default; see next section) — port 9876 is no longer hardcoded on the TCP data channel.
- Request:
[4 bytes big-endian length][JSON: {"id":"...", "command":"exact_exec", "expected":{"protocol_version":2,"instance_id":"...","pid":...,"project_path":"..."}, "request":{"script":"...","timeout":30}, "token":"..." (optional)}] - Response:
[4 bytes big-endian length][JSON: {"id":"...", "success":bool, "output":"...", "error":"...", "protocol_version":2, "instance_id":"...", "pid":..., "project_path":"..."}] - Token auth: required when the server binds non-loopback. The token is written to
<Project>/Saved/UnrealBridge/token.txt; clients read it and add"token":"<value>"to every request. Constant-time compared on the server. - Every request uses an
exact_*command with frozen discovery identity. Payload fields live underrequest, so a legacy server cannot mistakeexact_execfor an ordinary script request. The production dispatcher rejects missing/malformed identity, request objects, and unknown/legacy commands before any command body, work admission, or GameThread dispatch; the client rejects response identity drift.project_pathis one wire-canonical string and must match discovery/startup output exactly on every OS. - Client response frames are bounded to
MAX_RESPONSE_FRAME_BYTES(10 MiB), must be non-empty, valid UTF-8, and decode to a JSON object before identity fields are read. - Special exact commands (handled inline on the worker thread, bypass the Python exec queue):
exact_ping→pong(TCP-only liveness)exact_editor_status→ cached Engine/Slate tick ages, readiness, staleness and modal-attention summary (no fresh GameThread dispatch)exact_gamethread_ping→alive/unresponsive+latency_ms(GT liveness)exact_debug_resume→ unsticks a paused BP breakpoint viaFKismetDebugUtilities::RequestAbortingExecutionexact_modal_status→ structured active-Slate-modal snapshot (title, body, buttons, redacted inputs, checkboxes)exact_modal_action→ guarded click/input/checkbox action; rejects stale snapshots
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 · 145 lines · 3,525 tokens per session scan A 47795a55ede1
UnrealBridge CLAUDE.md is an instructions file published in the GitHub repository TornLux/UnrealBridge (300 stars, last pushed 2d ago), licensed MIT. It adds 3,525 tokens to every session, about $0.0176 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
claudemon CLAUDE.md
Instructions for zamarrowski/claudemon, covering ai agent guidelines, project shape, general guidelines, general rules and guards & defensive code.
unity-code-style-guide AGENTS.md
Instructions for krogh-jacobsen/unity-code-style-guide, covering agents.md — unity 6 c, project setup — edit this block, never do these — they corrupt the project, deprecated in unity 6 and if you read nothing else.
nes-php-glfw copilot-instructions.md
Instructions for oliverearl/nes-php-glfw, covering github copilot development guidelines for nes emulator, agent document source of truth, project overview, code style & standards and documentation.
game-and-watch-retro-go-sd CLAUDE.md
Claude Code instructions for sylverb/game-and-watch-retro-go-sd, covering claude.md, what this project is, build / flash workflow, architecture and three storage tiers, one elf.
base-building-trap-defense-design-agent-skill CLAUDE.md
Instructions for dungnotnull/base-building-trap-defense-design-agent-skill, covering claude.md — skill 250: base-building-trap-defense-design, skill identity, problem this skill solves, harness flow summary and sub-skills.
trident-mcp AGENTS.md
Instructions for mordor-forge/trident-mcp, covering agents.md, what this is, build and test, single-file verification and e2e tests.