datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
b3-agent-security-benchmark-weak[paper] [blogpost] [game]
b3 AI Security Benchmark: Breaking Agent Backbones
Highly contextalized prompt injections crowd-sourced during the Gandalf Agent Breaker Challenge.
This is a low-quality version of the data behind Breaking Agent Backbones: Evaluating the Security
of Backbone LLMs in AI Agents.
The high quality dataset was used to evaluate the security of more than 30 LLMs.
Dataset Summary
Purpose: This dataset contains crowdsourced adversarial attacks… See the full description on the dataset page: https://huggingface.co/datasets/Lakera/b3-agent-security-benchmark-weak.ai-agent-security-incidents
AI Agent Security Incident Database v0.1
A structured, machine-readable database of 1365 confirmed AI agent security incidents, collected and classified automatically.
What is this?
Every time an AI agent causes unintended harm — escaping a sandbox, exploiting an API, taking unauthorized actions, exfiltrating data — this database captures it.
This is not a list of theoretical risks. Every entry describes something that actually happened, with a verifiable source… See the full description on the dataset page: https://huggingface.co/datasets/gemmozero/ai-agent-security-incidents.nse-india-security-master
TickerTruth — NSE India Security Master (Explorer)
A clean, normalized reference table of 2,389 NSE-listed equities — ISIN mappings, listing dates, company names, and active/delisted status — built from the TickerTruth India reference-data pipeline.
Why this dataset exists
India equity data is notoriously messy. NSE symbols change (renames, mergers, delistings), ISINs get reissued, and raw bhavcopy files carry no historical context. TickerTruth's pipeline… See the full description on the dataset page: https://huggingface.co/datasets/tickertruthorg/nse-india-security-master.Security-TTP-Mapping
The Security Attack Pattern (TTP) Recognition or Mapping Task
We share in this repo the MITRE ATT&CK mapping datasets, with training, validation and test splits.
The datasets can be considered as an emerging and challenging multilabel classification NLP task, with over 600 hierarchical classes.
NOTE: due to their security nature, these datasets contain textual information about malware and other security aspects.
Datasets
TRAM
This dataset belongs to CTID… See the full description on the dataset page: https://huggingface.co/datasets/tumeteor/Security-TTP-Mapping.nse-india-security-master
TickerTruth — NSE India Security Master (Explorer)
A clean, normalized reference table of 2,389 NSE-listed equities — ISIN mappings, listing dates, company names, and active/delisted status — built from the TickerTruth India reference-data pipeline.
Why this dataset exists
India equity data is notoriously messy. NSE symbols change (renames, mergers, delistings), ISINs get reissued, and raw bhavcopy files carry no historical context. TickerTruth's pipeline… See the full description on the dataset page: https://huggingface.co/datasets/tickertruth/nse-india-security-master.Cyber-Security-Breachesbabbelphish
BabbelPhish
BabbelPhish is a dataset based on the Sublime Security Message Query Language (MQL) used for email security detection engineering. This dataset is specially created for the BabbelPhish project, which focuses on leveraging large language models to facilitate the work of detection engineers.
This dataset comprises around 3,000 examples drawn from various sources. We've utilized the following:
Sublime Security Documentation
Message Data Model (Schema)
Sublime Rules Repo… See the full description on the dataset page: https://huggingface.co/datasets/sublime-security/babbelphish.us-social-security-medicare-FAQs-testjapanchoice-2026-foreign-policy-and-securityExported: 2026-02-20
This dataset covers the topic of Foreign Policy and Security (外交・安全保障) from the Japan Choice Polis platform. The conversation can be found at: https://polis.japanchoice.jp/3x3ancmrtc
Data was exported from the database and provided by Nishio Hirokazu. It was gathered using the Polis software by The Computational Democracy Project (see: compdemocracy.org/polis and github.com/compdemocracy/polis).
US_Social_Security_Medicare_FAQs_Samplereal-world-benign-use-cases
Real-World Benign Use Cases
A curated set of 178 real-world, 100%-benign examples (label == 0 for every row) pulled from
production AI-coding-agent traffic — chat messages, tool output, shell commands, code snippets —
built specifically to stress-test prompt-injection / jailbreak classifiers for false positives.
Every row was independently judged benign with high confidence before inclusion. This is not a
random sample of production traffic: rows were preferentially drawn from… See the full description on the dataset page: https://huggingface.co/datasets/rogue-security/real-world-benign-use-cases.security_contentnetwork-security-route-hijack-coherence-risk-v0.1What this repo is for
Detect routing security incidents fast.
Covers:
unexpected origin ASN
invalid ROAs
AS-path anomalies
suspicious more-specific prefixes
observed traffic diversion
whether mitigation happened
This is high-impact because one leak can break many networks.
japanchoice-2025-foreign-policy-and-securityFetched: 2026-02-16-2245
Data was gathered using the Polis software (see: compdemocracy.org/polis and github.com/compdemocracy/polis), and acquired from this URL: https://polis.japanchoice.jp/7cdcyjmsyh
comments.csv was augmented with is-seed and is-meta columns fetched from https://polis.japanchoice.jp/api/v3/comments?conversation_id=7cdcyjmsyh&moderation=true&include_voting_patterns=true
openclaw-security-newsSource Repo: https://github.com/joylarkin/openclaw-security-newsSource Feed: https://raw.githubusercontent.com/joylarkin/openclaw-security-news/main/feed.xml
Description: OpenClaw Security News for AI builders, developers, and investors. Featuring global government warnings about OpenClaw, OpenClaw news headlines, and OpenClaw security vendor advisories.
Last Update: 22 May 2026
social_security_embeddingssecurityincidentszero_trust_network_security_logssecurity_steerability
Security Steerability & the VeganRibs Benchmark
Security steerability is defined as an LLM's ability to stick to the specific rules and boundaries set by a system prompt, particularly for content that isn't typically considered prohibited.
To evaluate this, we developed the VeganRibs benchmark. The benchmark tests an LLM's skill at handling conflicts by seeing if it can follow system-level instructions even when a user's input tries to contradict them.
VeganRibs works by presenting… See the full description on the dataset page: https://huggingface.co/datasets/itayhf/security_steerability.iot-securityIoT Data Collection
This document provides detailed information about the various data fields collected by the IoT Device with Multiple Sensors. Each field is described along with the type of data it includes.
Data Fields
ALS (Ambient Light Sensor)
Explanation: Measures the ambient light levels around the thermostat.
Data Type: Intensity of light in lux (lumens per square meter).
PIR (Passive Infrared Sensor)
Explanation: Detects motion by measuring the infrared (heat) emitted by objects in… See the full description on the dataset page: https://huggingface.co/datasets/fenar/iot-security.security_domain_knowlegesecurity_finetuneplaywright_security_shell_generationafrica-synth-urbanization-housing-tenure-security-africa-all
Africa Synth Urbanization Housing Tenure Security Africa All | Africa (Electric Sheep Africa metadata inventory)
Size category: 10K<n<100K - Formats: csv - Sector: governance_security - Engineered by Electric Sheep Africa
TL;DR
This dataset is part of the Electric Sheep Africa catalog on Hugging Face. It is indexed for African data discovery with standardized metadata, loading guidance, provenance notes, and analyst-oriented context.
What This… See the full description on the dataset page: https://huggingface.co/datasets/electricsheepafrica/africa-synth-urbanization-housing-tenure-security-africa-all.Human-Like-Reasoning-Security-Scenarios
Human-Like Reasoning Security Scenarios
A handcrafted dataset of realistic cybersecurity, system, and network scenarios
focused on human-style reasoning, not just final answers.
Why this dataset is different
Includes human thought processes
Covers common mistakes
Explains correct decisions
Designed for reasoning-based AI models
Domains
Cybersecurity
Systems
Networking
Use Cases
LLM fine-tuning
AI agent training
Cybersecurity education… See the full description on the dataset page: https://huggingface.co/datasets/Perfectyash/Human-Like-Reasoning-Security-Scenarios.viet-robot-security-events
viet-robot-security-events
A synthetic dataset of simple security-related events in a smart home.
Columns:
id: row id
event_type: door_open, window_open, motion_detected, noise_high, device_disconnected
location: where the event happened
time_hour: integer hour of day (0-23)
is_night: yes/no
human_confirmed: yes/no (whether a human was seen)
note: short English note
For demos only, not real security logs.
License
MIT
viet-robot-security-event-log
viet-robot-security-event-log
A small synthetic dataset of simple indoor security-related events observed
by a home robot.
Columns:
id: row id
timestamp_local: local time string
event_type: door_open, motion, sound, unknown
location: room or area
time_of_day: morning, afternoon, evening, night
severity: info, warning, critical
note_en: short English note
License
MIT
asss-securitysecurity-framework-crosswalk
Security Framework Control Crosswalk
197 CSA CCM v4 controls mapped to ISO/IEC 27001:2022, SOC 2 (Trust Services Criteria),
NIST CSF 2.0, NIST 800-53 Rev. 5, and Shared Assessments SIG — with per-mapping
confidence and provenance. Answer one framework, see what it covers in the others.
🔎 Interactive explorer: https://accountmade.com/tools/framework-crosswalk
💻 Source repo: https://github.com/accountmade/security-framework-crosswalk
942 mappings across 197 controls / 17 CCM… See the full description on the dataset page: https://huggingface.co/datasets/jakexkim/security-framework-crosswalk.security_analysis