Навыки: Безопасность

Detects LLM prompts built from user input or retrieved documents that let attackers hijack the model's instructions. Use...

确定性检测文档、网页片段和工具结果中的指令覆盖、角色伪造、工具诱导、数据外传、安全绕过及隐藏编码特征，适用于不可信内容进入模型前的安全检查。

Generate, audit, tighten, adapt, or decompile prompts for the GPT-5.6 model family. Use when the user wants a first-pass...

Run a phased terseness pass over an existing prompt or skill and return a diff-proposal: mechanical wording and format c...

Prompt injection testing. USE WHEN prompt injection, jailbreak, LLM security, AI security assessment, pentest AI applica...

Use when verifying ZK proofs server-side, validating proofs before transaction submission, building verification gateway...

Develop evidence-backed structural and architectural security hardening proposals from vulnerability disclosures, suppli...

Post-hoc verification and trust audit of AI-generated factual and interpretive claims. Classifies claims by type and sal...

Agent governance skill for MCP tool calls — Cedar policy authoring, shadow-to-enforce rollout, and Ed25519 receipt verif...

Configure Cedar policy enforcement and Ed25519 signed receipts for Claude Code tool calls. Use when setting up projects...

Super-Admin-only lock-down pattern. Use before touching admin.admins, admin.admin_audit_log, is_super_admin() / requireS...

Assesses network protocol exploitation candidates while separating observed service behavior, advisory applicability, re...

Master network protocol reverse engineering including packet analysis, protocol dissection, and custom protocol document...

Reverse engineers network protocols through packet analysis, dissection, and documentation. Use when analyzing network t...

Transport security where the client establishes trust: TLS version floor, certificate chain and hostname verification, t...

Structured, protocol-type-specific audit templates enumerating the exact checks, invariants, and attack vectors relevant...

Detects JavaScript deep-merges or property assignment from user input that can pollute Object.prototype. Use when writin...

Find what a change could break beyond the diff, then prove the one fact it is safe because of by running real code. Appl...

Extends power/interest stakeholder mapping with the actor types specific to public-sector projects (elected officials, c...

Conduct rigorous multi-perspective diff interrogation and code review. Categorize all findings strictly into 'act on', '...

Keep a concise, append-only decision and evidence trail for long or multi-phase work. Use when the user needs to audit c...

Maintain local pstack adaptations against the official upstream: compare versions, apply useful compatible updates, and...

publication-link-audit — Verify every outbound URL in a Reddit post, megathread, guide, wiki page, or README before publ...

Assess and route one bounded Proto UI dependency drift, advisory, or update question. Use to trace manifest and lockfile...

Audit one bounded Proto UI collaboration-governance slice and route policy-determined drift toward reversible correction...

Audit an already completed protected Proto UI publication and reconcile immutable external facts with Version entities,...

Reviews a pull request or other code change for correctness bugs, security flaws (OWASP Top 10, STRIDE), performance reg...

Unabhängiges Code-Review der Pulse-Änderungen (Branch gegen main oder ein PR) vor dem Zusammenführen – in einem frischen...

Audit any website for what makes it look AI-generated or unfinished, such as dead buttons, missing hover and focus state...

Send a push notification via ntfy.sh or Pushover

Drive PWN::AWS::STS from pwn_eval.

Full-chain engineering methodology from reverse engineering to working exploit. Applicable scenarios: got binary + vulne...

Drive PWN::Plugins::SSN from pwn_eval.

Drive PWN::Reports::XML from pwn_eval.

Drive PWN::SAST::AWS from pwn_eval.

Drive PWN::SAST::CSRF from pwn_eval.

Drive PWN::SAST::Eval from pwn_eval.

Drive PWN::SAST::Log4J from pwn_eval.

Drive PWN::SAST::Logger from pwn_eval.

Drive PWN::SAST::MD5 from pwn_eval.

Drive PWN::SAST::Port from pwn_eval.

Drive PWN::SAST::ReDOS from pwn_eval.

Drive PWN::SAST::Shell from pwn_eval.

Drive PWN::SAST::SQL from pwn_eval.

Drive PWN::SAST::SSL from pwn_eval.

Drive PWN::SAST::Sudo from pwn_eval.

Drive PWN::SAST::Token from pwn_eval.

Official Qt for Python (PySide6) development, QML integration, asynchronous signals, and UI compilation.