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/gabrielmaialva33/camera-hack/claude-mdgit clone --depth 1 https://github.com/gabrielmaialva33/camera-hackWrote 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/gabrielmaialva33/camera-hack/claude-md)<a href="https://agentmods.dev/instructions/gabrielmaialva33/camera-hack/claude-md"><img src="https://agentmods.dev/badge/instructions/gabrielmaialva33/camera-hack/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 | $0.00468 | $0.00468 |
| Opus 5 | $0.00234 | $0.00234 |
| Sonnet 5 | $0.00094 | $0.00094 |
| Haiku 4.5 | $0.00047 | $0.00047 |
Grade B, and why
camera-hack CLAUDE.md scanned grade B with 1 finding 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.
Asks for rootmediumPrivilege escalation
A mod that escalates privileges can change anything on the machine, not only the project.
Ports only respond to SYN scans (`sudo nmap -sS`), not TCP connect. What it actually says
CLAUDE.md
Project
Security research on Yoosee/Gwelltimes IP cameras. Current target: SC-B21 dual-lens PTZ (firmware 38.02.103).
Attack Surface
| Port | Service | Auth | Notes |
|---|---|---|---|
| 554/tcp | RTSP | Partial | /onvif1 needs auth, /cam/realmonitor open |
| 5000/tcp | ONVIF | None | Full enumeration, PTZ control, DoS via write ops |
Ports only respond to SYN scans (sudo nmap -sS), not TCP connect.
RTSP Streams
| Path | Resolution | Auth |
|---|---|---|
/onvif1 |
1920x2160 H.265 15fps | Required |
/onvif2 |
640x720 H.265 15fps | Required |
/cam/realmonitor |
Varies | None |
Structure
├── advisories/ # Formal vulnerability writeups
├── tools/rtsp/ # RTSP capture tools
├── sdcard_hack/ # SD card exploitation
├── arduino/ # UART serial bridge
├── esp32/ # ESP32 serial bridge
└── config.example # Config template (no credentials)
Hardware
- OEM: Shenzhen Gwelltimes Technology
- Firmware: 38.02.103, Hardware Ver 2.1
- Video: Dual lens stitched vertically (1920x2160 + 640x720)
- WiFi: RTL8188 variant
- Audio: G.711 PCM 16kHz mono
- MAC OUI:
4C:B0:08(Gwelltimes),A8:B5:8E(Bilian/WiFi)
Notes
- ONVIF write operations (CreateUsers, SetHostname) crash the IPC process permanently
- Camera requires power cycle to recover from DoS
- Scripts have placeholder IPs — copy
config.exampletoconfig.env
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.
- 3d ago First seen · 49 lines · 468 tokens per session scan B 58d4bea984bc
camera-hack CLAUDE.md is an instructions file published in the GitHub repository gabrielmaialva33/camera-hack (21 stars, last pushed 4mo ago), licensed MIT. It adds 468 tokens to every session, about $0.0023 per session on Opus 5. A static security scan graded it B with 1 finding (asks for root). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other instructions, from other repositories
FastLED GEMINI.md
Gemini CLI instructions for FastLED/FastLED, a project described as: The FastLED library for colored LED animation on Arduino. Please direct questions/requests for help to the FastLED Reddit community: http://fastled.io/r We'd like to use github "issues" just for tracking library bugs / enhancements.
jlink-mcp CLAUDE.md
Instructions for Klievan/jlink-mcp, covering j-link mcp server, building, testing, vscode extension (native mcp integration) and standalone mcp server (for claude desktop / claude code).
platform-nordicnrf54 AGENTS.md
Instructions for caveman99/platform-nordicnrf54, covering agent guide for this repo, what this repo is, companion repos, project conventions and commit messages.
serial-mcp CLAUDE.md
Instructions for qarnet/serial-mcp, a project described as: Serial port MCP server for Claude Code, Codex, Opencode and AI agents. Reliable UART/USB-serial access for embedded development.
FastLED CLAUDE.md
Claude Code instructions for FastLED/FastLED, covering fastled ai agent guidelines, read the right file for your task, key commands, core rules (all agents) and git and code publishing.
ROS AGENTS.md
Instructions for lawliet206/ROS, covering agents.md, 项目是什么, 环境版本(不可随意升级/降级), 目录结构(catkin 工作空间根 = 仓库根) and 重要 launch 与脚本(命名与参数以代码为准,改文档必须同步).