Repository navigation
Add offensive-prompt-injection skill (LLM prompt injection & jailbreaks) - #45
Open
rudycelekli wants to merge 2 commits into
Open
rudycelekli wants to merge 2 commits into
rudycelekli wants to merge 2 commits into
Conversation
offensive-prompt-injection under ai; complementary to offensive-ai-security. Direct + indirect/second-order injection, jailbreak families, agentic tool abuse, multimodal, guardrail bypass; assessment harness (garak/PyRIT/promptmap/Giskard); OWASP LLM Top 10 + MITRE ATLAS. Updates README, CHANGELOG, claude-skills.json.
build_manifest.py counts 80 skills because main already contains offensive-netexec, which the published badge and manifest omit. The badge now matches skill_count.
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Adds
offensive-prompt-injectionunderai— a dedicated prompt-injection surface, complementary to the broaderoffensive-ai-securityskill.Coverage: direct and indirect/second-order injection (RAG docs, web pages, email, file metadata, alt text/OCR); system-prompt/instruction extraction; jailbreak families (persona, payload splitting, base64/rot13/leetspeak/homoglyph/zero-width obfuscation, token smuggling, many-shot, Crescendo, refusal suppression, context overflow); agentic tool abuse (markdown-image/link exfil beaconing, SSRF, function-arg injection, confused-deputy, memory poisoning); multimodal injection; and guardrail/classifier bypass. Includes a canary-based assessment harness with garak/PyRIT/promptmap/Giskard examples and a defender view (spotlighting, least-privilege tools, dual-LLM, HITL, allow-listing).
Follows the SKILL.md format with Key References mapped to OWASP Top 10 for LLM Applications (LLM01/LLM02/LLM06) and MITRE ATLAS. README, CHANGELOG, and
claude-skills.jsonupdated viatools/build_manifest.py.Made with Cursor