prompt-injection-tester

prompt-injection-tester is a skill for Claude Code from NovaCode37/claude-security-skills. It costs 84 tokens per session (904 once invoked), scanned A, original, MIT.

A defensive test tool for checking whether an AI application resists prompt injection and jailbreak attempts. Prompt injection is an attempt to make an AI ignore its rules, reveal hidden instructions, or expose data.

In plain words
What is it for?
Use it to test instruction overrides, system-prompt leaks, role-play attacks, delimiter escapes, encoded attacks, indirect data theft, and refusal bypasses, including in CI after prompt changes.
Why use it?
It provides repeatable attack tests and reports whether the application's safeguards held up. It is intended for systems you own or are authorized to test.

Skill for Claude Code

Written for Claude Code: shipped in a Claude Code plugin.

Part of the claude-security-skills plugin — 8 skills shipped together

Good fit Use it to test instruction overrides, system-prompt leaks, role-play attacks, delimiter escapes, encoded attacks, indirect data theft, and refusal bypasses, including in CI after prompt changes.

Compare 6 skills from other repositories ↓
Install with agentmods
npx agentmods add skills/novacode37/claude-security-skills/prompt-injection-tester
Install

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.

Any agent
npx skills add NovaCode37/claude-security-skills --skill prompt-injection-tester
Clone the repo
git clone --depth 1 https://github.com/NovaCode37/claude-security-skills

Made for: Claude Code.

Or install claude-security-skills, the plugin that ships this one along with the rest of its 8 skills.

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 prompt-injection-tester

README.md
[![agentmods](https://agentmods.dev/badge/skills/novacode37/claude-security-skills/prompt-injection-tester/github.svg)](https://agentmods.dev/skills/novacode37/claude-security-skills/prompt-injection-tester)
Your own site
<a href="https://agentmods.dev/skills/novacode37/claude-security-skills/prompt-injection-tester"><img src="https://agentmods.dev/badge/skills/novacode37/claude-security-skills/prompt-injection-tester/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 prompt-injection-tester

Your own site · 80×15
<a href="https://agentmods.dev/skills/novacode37/claude-security-skills/prompt-injection-tester"><img src="https://agentmods.dev/badge/skills/novacode37/claude-security-skills/prompt-injection-tester.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 84 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 904 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe. Third-party audits
  • NVIDIA SkillSpector pass 7 Sept 2026
How audits are shown
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 $0.00084 $0.00904
Opus 5 $0.00042 $0.00452
Sonnet 5 $0.00017 $0.00181
Haiku 4.5 $0.00008 $0.00090

Measured 2d ago against content hash 9cee101b91ad, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-12, from the pricing page.

Security

Grade A, and why

prompt-injection-tester 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 2d ago.

The scan reads SKILL.md. This mod also ships 2 executable files (attacker.py, tests/test_attacker.py), listed below but not scanned — reading those needs a real analyzer, not pattern matching.

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.

skills/prompt-injection-tester/SKILL.md · 105 lines

How it starts

The opening of the file, as written. The whole thing — 105 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Prompt Injection Tester

A defensive red-team harness for evaluating the prompt-injection resistance of LLM applications you own or are authorized to test. It ships a library of well-documented public attack techniques and a canary-based detection engine that decides whether each attack succeeded — then scores overall resilience.

⚠️ Use only against systems you own or have permission to test. The payloads are public hardening techniques, intended to strengthen guardrails.

When to use this skill

  • "Is my chatbot vulnerable to prompt injection / jailbreaks?"
  • "Red-team / pentest my LLM app's system prompt."
  • "Score how well my guardrails resist instruction-override attacks."
  • Regression-testing guardrails in CI after a prompt change.

Attack categories covered

instruction-override · system-prompt-leak · role-play (DAN-style) · delimiter-escape · encoding (base64/leetspeak) · data-exfiltration (indirect injection) · refusal-suppression.

How it works

  1. A unique canary secret is embedded into a guarded system prompt.
  2. Each payload is sent as the user turn.
  3. The response is scored: it's vulnerable if it hits an attack success-marker or leaks the canary; resisted if it refuses.
  4. You get a resilience score (0–100) and a per-category breakdown.

How to run it

List the payload library (no model calls):

python skills/prompt-injection-tester/attacker.py --list
python skills/prompt-injection-tester/attacker.py --list --category role-play

See the harness detect a weak vs. a hardened model:

python skills/prompt-injection-tester/attacker.py --demo            # weak model
python skills/prompt-injection-tester/attacker.py --demo --hardened # safe model
python skills/prompt-injection-tester/attacker.py --demo --json

Exit codes: 0 clean (no payload succeeded, or --list) · 1 at least one payload succeeded · 2 usage error.

Read the full file on GitHub · 105 lines

Files

What ships with it

2 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.

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. 2d ago Changed · +3 lines 9cee101b91ad
  2. 12d ago First seen · 102 lines · 84 tokens per session scan A d91d947b583a

Subscribe to this mod's changes

prompt-injection-tester is a skill published in the GitHub repository NovaCode37/claude-security-skills (11 stars, last pushed 3d ago), licensed MIT. It adds 84 tokens to every session and 904 once invoked, about $0.0004 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.

Related

Other skills, from other repositories

qa-testing

Run QA testing on a page, feature, or full site at one of three depth tiers (smoke, standard, full). Use this skill whenever the user asks to QA a page or site, run a smoke test after a deploy, verify a page before launch, or run a regression sweep. Triggers on QA, QA sweep, smoke test, regression test, post-deploy…

rampstackco/claude-skills · 170 tokens

visual-qa

Runs rigorous visual QA across web, terminal, and paginated surfaces with screenshot evidence and a verdict. Use for any UI build or change, or when asked whether a page, component, or TUI looks right.

code-yeongyu/oh-my-openagent · 47 tokens

remove-ai-slops

Removes AI-generated code smells from branch changes or an explicit file list behind regression tests. Use when the user asks to clean up, deslop, or remove AI-slop patterns from recent changes.

code-yeongyu/oh-my-openagent · 45 tokens

agent-browser

Automates browser interactions for web testing, form filling, screenshots, and data extraction. Use when the user needs to navigate websites, interact with web pages, fill forms, take screenshots, test web applications, or extract information from web pages.

code-yeongyu/oh-my-openagent · 51 tokens

senpi-qa

QA the omo Senpi adapter (packages/omo-senpi, packages/senpi-task) against the REAL senpi binary in strict isolation, and write every artifact to the one canonical evidence path .omo/evidence/omo-senpi-adapter/ /. The live drivers under packages/omo-senpi/scripts/qa/ create their own isolated SENPICODINGAGENTDIR and…

code-yeongyu/oh-my-openagent · 251 tokens

suede-ai-eval

Suede Labs AI eval design and coverage audit: AI-SPEC, failure-mode rubric with severity scoring, concrete pass/fail eval cases, coverage and infrastructure scores, and mechanical acceptance gates. Use when a change ships LLM, RAG, agent, classifier, prompt, or generated-media behavior, or when asked to write evals…

JasonColapietro/suede-creator-skills · 178 tokens