GraphCanon updated Oct 5, 2026 · GitHub synced Oct 5, 2026
33views this month
Decision brief
PromptAttack is an LLM-targeted adversarial attack tool that leverages prompt engineering to generate adversarial samples keeping semantic intact but misclassifying outputs.
Good fit when
- For targeted analysis of adversarial robustness in specific language models.
- When needing a detailed understanding of how perturbations in input prompts affect model classification performance.
Avoid when
- If the focus is on general model improvement rather than adversarial testing.
- When working with proprietary or sensitive data that cannot be manipulated via external prompt tools, given potential data leakage concerns.
Observed Jul 17, 2026 · Source: enrich:decision_facts
Verify the decision
Maintenance and security
Full trust report- Maintenance
- Dormant (622d since push)
- As of Oct 5, 2026
- Provenance
- Not a fork · Personal account
- As of Oct 5, 2026
- Security (OSV)
- 241 low (241 low)
- As of Jul 11, 2026
Public GitHub metadata and optional OSV scans. Signals, not a guarantee. Trust methodology.
Install
pip install PromptAttackSimilar tools
Same-category neighbours. No typed graph edges are catalogued for this tool yet.
Evidence and technical details
Sourced facts, taxonomy, compatibility claims, README excerpt, and machine-readable endpoints.
Overview
This repository contains the source code for the ICLR 2024 paper that discusses a prompt-based adversarial attack on language models, allowing users to generate adversarial samples targeting various LLMs.
Capability facts
- Languages
- python
Source: github.language · Oct 5, 2026
Categories
Compatibility
Sourced claims from the README excerpt - not unsourced marketing copy.
Source: README excerpt (regex_v1, Oct 5, 2026)
pip install -r requirements.txtSource link
Source: README excerpt (regex_v1, Oct 5, 2026)
Logs of querying *ChatGPT*:Source link
Tags
README
An LLM can Fool Itself: A Prompt Based Adversarial Attack This is the source code for the ICLR 2024 paper "An LLM can Fool Itself: A Prompt Based Adversarial Attack", <br Xilie Xu (NUS), Keyi Kong (SDU), Ning Liu (SDU), Lizhen Cui (SDU), Di Wang (KAUST), Jingfeng Zhang (Universit...
For agents
This page has a .md twin and JSON over the API.