Cells for security work: scanners, secret stores, sandboxes, guards, audits, pentest tools and the directories of the field's forums and markets.

1. model-tiersby elder-plinius (the leak), Nodecode (this cell) · rewrite of elder-plinius/CL4R1T4SThe Claude model hierarchy as the leaked Claude Code prompt tells it: the Mythos tier above Opus, with line numbers53kCyber security
2. jailbreaksby togg53192-cmd (the collection), Nodecode (this cell) · rewrite of togg53192-cmd/jailbreaksA working collection of jailbreak prompts, indexed: which model each one targets, what kind it is, and any of them quoted by line17kCyber security
3. t3mp3stby Pliny the Elder · rewrite of elder-plinius/T3MP3STDrive a running T3MP3ST War Room: launch a mission, watch it, read findings7.1kCyber security
4. redteam-gatesby SeaOf0 · rewrite of SeaOf0/dsh-redteam-modelRed-team stage gates checked by a tool: what a stage's artifacts hold, and what is missing879Cyber security
5. tracekitby Cygnux-Labs · rewrite of Cygnux-Labs/TracekitTracekit: a signer outside the agent decides every tool call and signs the run726Cyber security
6. bubblewrapby NousResearch · rewrite of NousResearch/hermes-official-pluginsBubblewrap: run a command in its own bwrap sandbox, read-only host, secrets hidden725Cyber security
7. security-scanby NousResearch · rewrite of NousResearch/hermes-official-pluginsSecurity reviews of a local repository: nine tools, /security, and md/json/sarif/csv reports725Cyber security
8. snykby NousResearch · rewrite of NousResearch/hermes-official-pluginsSnyk vulnerability scans of code, dependencies, containers and IaC, as tools the model calls725Cyber security
9. deepsecby Unclecheng-li · rewrite of Unclecheng-li/DeepSecA local security review of a tree: secrets, risky code, agent-config and supply-chain checks, as tools517Cyber security
10. deepcanaryby Oscar-Williams · rewrite of Oscar-Williams/dsh-deepcanaryAn attention inbox with levels and quiet hours for the operator174Cyber security
11. pentesterby fb0sh · rewrite of fb0sh/dsh-pentesterPTES pentest run: seven stages, nineteen specialists, deliverables on disk102Cyber security
12. secure-auditby PensiveFei · rewrite of PensiveFei/dsh-secure-auditPrompt-injection scan, Chinese-PII redaction and a local configuration audit86Cyber security
13. blackboxby umanitek · rewrite of umanitek/agent-blackboxAgent Blackbox, local half: scan untrusted text and shell commands for threats69Cyber security
14. upstash-boxby upstash · rewrite of upstash/boxUpstash Box: run a shell command, read and write files in a cloud Linux container45Cyber security
15. little-canaryby hermes-labs-ai · rewrite of hermes-labs-ai/little-canaryScreens untrusted text for prompt injection with a local canary model, and withholds a session's tools on a BLOCK41Cyber security
16. osint-mapby Adolanium · rewrite of Adolanium/hermes-osint-mapOSINT Map: public situational feeds and passive recon as five tools, with no key32Cyber security
17. defendby PerryLink · rewrite of PerryLink/dsh-defendInjection, jailbreak and secret scanning: scans a tool call's text before it runs, with allow, ask and block tiers28Cyber security
18. enhance-scopeby wufeng0105 · rewrite of wufeng0105/dsh-enhanceEnhance: your text at the end of every system prompt, the opening line rewritten, and a tool veto23Cyber security
19. plugin-vetby PerryLink · rewrite of PerryLink/dsh-skill-pack-securityPlugin Vet: a five-dimension supply-chain risk card for a repository, npm package or folder22Cyber security
20. maskby PerryLink · rewrite of PerryLink/dsh-maskMasks phones, emails, ID cards, bank cards and keys in the text you send17Cyber security