~ / leaderboard / skill / prompt-injection-defense
Skill · #2321

prompt-injection-defense

Defend the agent against adversarial Unicode it obeys but a human can't see: bidi/Trojan Source, tag-char ASCII smuggling, zero-width text, forged line breaks, terminal escapes, homoglyphs. Use when a

7rigs use it
0%of all rigs
5.3kavg rig context tax

Get it

$ npx degit DevelopersGlobal/ai-agent-skills/skills/prompt-injection-defense .claude/skills/prompt-injection-defense

From the most-starred rig that ships it: DevelopersGlobal/ai-agent-skills. Same-named skills in different rigs may differ — review before use.

Often used together with

Rigs using prompt-injection-defense (7)

DevelopersGlobal/ai-agent-skills
AI agent skills for production grade applications
★ 66 collection Claude CodeCodexGemini CLI 31 skills 3 agents 7mo ago
Skill Collector 2.0k tok · · Q0
0xDarkMatter/claude-mods
Expert skills, agents, commands, rules, hooks & output styles for Claude Code — session continuity + modern CLI tooling for real-world dev workflows
★ 62 collection Claude CodeCodex 112 skills 3 agents 1mo ago
Skill Collector 10.4k tok · · Q3
radif-ru/ai-multi-agent-system
Локальная мульти-агентная система на self-hosted LLM (Ollama), 0 обращений к облачным LLM-API. Роли Planner/Executor/Critic, цикл thought → action → observation. 3 канала: Telegram, MAX, консоль. 19 инструментов: почта, Яндекс.Диск, веб, OC
★ 6 project Claude CodeCodexCursor 24 skills 1mo ago
Skill Collector 3.1k tok · · Q0
XData-si/xdata-agenthub
Open-source, vendor-neutral hub for discovering, reviewing and installing reusable AI agent skills, instructions and workflows.
★ 1 collection Claude CodeCodex 104 skills 1 agents 12d ago
Skill Collector 8.2k tok · · Q10
uqeu/estelle-cli
The trust layer under your AI coding agents. Estelle grounds every model in your real codebase and refuses to ship what isn't true: repository memory, a deterministic grounding gate, and grounded review. Model-agnostic, BYOK.
★ 0 project Codex 26 skills 25d ago
Skill Collector 1.4k tok · · Q0
consciontologic/agentic-workspace
A model-agnostic scaffold for repositories where AI coding agents do real work. Drop-in instructions, configs, skills, tracking, and a tiny Makefile + xops layer
★ 0 project Claude CodeCodexCopilot 1 mcp 45 skills 27d ago
Skill Collector 8.4k tok · · Q8
Mariano215/tack
Run a different Claude Code and Codex setup in every terminal. One profile drives both agents, with per-terminal config isolation.
★ 0 setup Claude CodeCodex 30 skills 3 agents 1mo ago
Skill Collector 3.9k tok · · Q5
copied ✓