「kernel」の検索結果

Endpoint defense evasion after an operator-selected chain already has a valid execution path but AV/EDR/AMSI/WDAC/PPL/memory/kernel telemetry blocks the intended action. Originating modules include /web-attack /ad-attack /cloud-attack /k8s /cicd /service-attack /phishing /privesc-win /privesc-linux /creds /post /shell. Success is the blocked action becoming executable, then resume the originating module — not obtaining a shell here.

日本語の概要は準備中です。原文の説明を表示しています。

Windows local privilege escalation from a low-privilege shell to Administrator or SYSTEM. Covers quiet vs loud enumeration, SeImpersonate/Potato family including LocalPotato, token privileges (SeBackup/SeRestore/SeManageVolume/SeLoadDriver/SeDebug), service/DLL/COM/scheduled-task abuse, AlwaysInstallElevated and UAC boundary, KrbRelayUp-style domain-joined local admin, and version-gated kernel LPE. Use when the operator already has a Windows foothold and needs local SYSTEM. Endpoint blocks hand off to /edr-bypass then return here.

日本語の概要は準備中です。原文の説明を表示しています。

Linux local privilege escalation from a low-privilege shell to root. Covers quiet vs loud enumeration, sudo/GTFOBins, CVE-2025-32463 chwoot and CVE-2025-32462 host bypass, polkit/udisks CVE-2025-6018/6019, SUID/capabilities, systemd timers/units, cron/PATH/LD_PRELOAD, dangerous groups and docker.sock, host-local container escape, and version-gated kernel LPE including Copy Fail CVE-2026-31431. Use when the operator has a Linux foothold and needs root. Kubernetes RBAC stays in /k8s. Endpoint blocks hand off to /edr-bypass then return here.

日本語の概要は準備中です。原文の説明を表示しています。

Use when verifying that a model in the imp inference engine produces coherent output without repetition loops, token-stuck states, or state corruption across turns or streams. Triggers on "degenerates", "check degeneration", "repetition loop", "own own own", "stuck token", "empty content", "multi-turn regression", "does it still work", "NIAH", and after enabling CUDA graphs / changing forward pass / MoE routing / KV cache or KV dtype / GDN state or scan / sparse attention / PDL / speculation (MTP, n-gram) / ragged prefill / batched-decode kernels (smallm, producer quantize) / FA2 softmax.

日本語の概要は準備中です。原文の説明を表示しています。

Use when imp outputs differ between two paths that should agree - chunk sizes, prefill vs decode, prefix-cache resend, two engines (imp vs llama.cpp vs HF), two builds or kernels - "output changes with chunk size", "first token differs", "logprobs moved", "which layer diverges", "who is right, imp or llama.cpp", "near-tie". Do NOT use for repetition loops or garbage output (check-degeneration), a new arch that loads wrong (add-model-arch), or perf (benchmark-cuda).

日本語の概要は準備中です。原文の説明を表示しています。

Use when auditing the imp codebase for structural debt, dead code, god-objects/files, duplication, flag sprawl, or deciding whether a cleanup is worth shipping - "structure audit", "tech debt", "is this still used", "dead code", "refactor for clarity", "should we split this file", "remove this flag", "file size gate red", "alloc sites gate red". Do NOT use for build/test mechanics (building-and-testing), perf/kernel work (benchmark-cuda / sm120-cuda-expert), or output quality (check-degeneration).

日本語の概要は準備中です。原文の説明を表示しています。

Use when writing, moving or auditing any .md in imp - deciding which file a paragraph belongs in, adding a doc, fixing a stale claim, or when docs_lint.py fails in CI. Covers the four reader layers (L0 README / L1 operators / L2 kernel devs / L3 agents), the HTML-comment metadata header, [PROV:] provenance, the single-source-of-truth map, generated perf blocks, which numbers may appear in prose, plan-doc closure. Triggers on "which doc does this go in", "docs lint failed", "add a doc", "this claim is stale", "update the README", "PROV block", "layer", "STALE.md". Do NOT use for the CHANGELOG or a release body (shipping-prs), or for code-comment accuracy (codebase-audit).

日本語の概要は準備中です。原文の説明を表示しています。

Use when working on or testing imp-server and its HTTP APIs - OpenAI/Anthropic endpoints, /v1/chat/completions, /v1/messages, /v1/responses, SSE streaming, tool calling, json_schema/regex/GBNF constrained decoding, thinking/reasoning_content, cache_control/prefix cache, model loading and swapping, request priority, X-Request-Id, OTLP tracing / traceparent / Jaeger, /metrics, container env (IMP_SET), "server returns 404/garbage", API compliance. Do NOT use for CLI-only inference (imp-cli), kernel work (sm120-cuda-expert), or pure output-quality checks (check-degeneration).

日本語の概要は準備中です。原文の説明を表示しています。

Use when adding support for a new model architecture to imp, porting a model family, or debugging a model that loads but produces wrong output - "add support for <model>", "new arch", loader detection, chat template, tokenizer parity, RoPE variant, "outputs garbage", "prompt-blind", "digits scrambled", "NaN logits", "describes a different picture", "does it fit in VRAM". Do NOT use for kernel performance (sm120-cuda-expert) or quant-format questions (quant-formats).

日本語の概要は準備中です。原文の説明を表示しています。

Use when writing eBPF programs with libbpf, bpftrace, or Rust with aya-ebpf and aya-log, attaching kprobes, tracepoints, or XDP hooks, declaring maps, sharing data with a Tokio loader, triaging verifier or Aya load errors, or porting with CO-RE. Not for full kernel bypass: use dpdk.

日本語の概要は準備中です。原文の説明を表示しています。

Use when creating AF_XDP sockets, configuring UMEM and XSK rings, writing an XDP redirect program, or choosing copy versus zero-copy mode. Not for full kernel bypass: use dpdk.

日本語の概要は準備中です。原文の説明を表示しています。

Consolidated Symfony conventions (Symfony 6.4 / 7.x): attribute routing, controllers as services, autowiring/DI, Form types, Validation constraints, Voters for authorization, the Serializer/DTO contract, and Messenger. Apply when writing or reviewing Symfony backend code. Activated automatically by symfony-plugin/stack.md as a convention skill for the development phase. Apply when: writing or reviewing Symfony controllers, services, forms, validators, voters, and serialization. Do NOT use this skill for: - PHP language idioms (strict_types, enums, readonly) — see php-foundation:php-conventions. - Doctrine queries / entities — see symfony-plugin:doctrine-patterns. - Testing — injected by the QA phase (WebTestCase/KernelTestCase) + php-foundation:php-testing.

日本語の概要は準備中です。原文の説明を表示しています。

PHPUnit and Pest patterns for any PHP project: test structure, naming, data providers, test doubles (mocks/stubs/fakes), fixtures, and coverage targets. Stack-agnostic — referenced by every PHP plugin in the marketplace. Use this skill to: - Structure unit and integration tests with PHPUnit or Pest consistently. - Drive cases with data providers / datasets instead of copy-pasted test bodies. - Use test doubles with discipline — mock collaborators, not the system under test. - Set up and tear down fixtures cleanly and aim for a meaningful coverage target. Do NOT use this skill for: - PHP language idioms — see php-foundation:php-conventions. - Composer / autoloading — see php-foundation:composer-tooling. - Framework-specific test helpers (Laravel RefreshDatabase/actingAs, Symfony WebTestCase/KernelTestCase) — those are injected by the framework plugin's QA phase.

日本語の概要は準備中です。原文の説明を表示しています。

Apply the Minimum Sufficient Work (MSW) principle through the MSW Kernel to scope, execute, verify, and stop agent work. State the requested outcome and smallest proof, admit a claim only when deleting it would leave the contract unmet or unproven, do and prove each necessary claim, and halt at the fixed point. Use for implementation, debugging, review, research, planning, documentation, or any task where scope control, proportional evidence, or avoiding overwork matters. Also use when the user says MSW, Minimum Sufficient Work, do only what is necessary, avoid over-engineering, or stop when the result is proven.

日本語の概要は準備中です。原文の説明を表示しています。

Apply the Minimum Sufficient Language (MSL) principle through the MSL Kernel to write anything a reader must act on. Bind the reader and what they already know, partition facts from the machinery that produced them, emit each admitted fact as an action, a verification, a judgment, or a state, and stop when every claim is checkable and the reader can act. Use for status updates, work-in-progress reports to your own user, closeout and convergence comments, tickets, pull request and review write-ups, incident and postmortem notes, emails, memos, and any draft for readers outside the work. Also use when the user says MSL, Minimum Sufficient Language, write this for stakeholders, drop the jargon, make this readable, or asks for a paste-ready draft.

日本語の概要は準備中です。原文の説明を表示しています。

当用户直接要求进行数据集压缩测评，或在量化调优过程中需要使用 AISBench 评测工具输出数据集精度等场景时使用。 本 Skill 负责：使用 AISBench 数据集压缩代码、用 RBF Kernel Herding 算法生成压缩后的数据子集（coreset），并按用户目标交付两种产物之一： ① 接入量化调优主流程：产出独立的 coreset 数据集配置（不覆盖原始数据），回传「全集 config_name + coreset config_name」给编排层，供 evaluation.yaml 切换； ② 仅数据集压缩测评（不做量化调优）：基于 example 模板生成「全集测试脚本」和「子集测试脚本」两份 shell 脚本。 暂时支持 AIME 2025 和 GPQA，其他数据集需要再结合 aisbench tools目录下 herding_coreset_selector 源码进行分析。

日本語の概要は準備中です。原文の説明を表示しています。

Use when the user asks about NixOS, Nix language, flakes, nixpkgs, configuration.nix, nixos-rebuild, modules, options, overlays, derivations, mkDerivation, callPackage, home-manager, nix-darwin, devenv, packaging, cross-compilation, binary caches, garbage collection, NixOS containers, networking, firewall, systemd services, filesystems, LUKS, Wayland, X11, kernel, profiles, NixOS testing or VM tests, ISO images, Docker images, Raspberry Pi, Terraform, distributed builds, remote machines, pinning nixpkgs, rollback, generations, Nix Pills, or any topic related to configuring, packaging, deploying, or troubleshooting NixOS and the Nix ecosystem.

日本語の概要は準備中です。原文の説明を表示しています。

Perform forensic analysis of the macOS Unified Logging System (ULS) to investigate system events, application crashes, kernel panics, and potential indicators of compromise (IoCs) such as persistence mechanisms or unauthorized access.

日本語の概要は準備中です。原文の説明を表示しています。

Identify and exploit misconfigured Linux Capabilities. This skill covers how attackers escalate privileges to root without relying on SUID binaries or kernel exploits by abusing excessive capabilities like cap_dac_read_search, cap_sys_ptrace, or cap_setuid assigned to ordinary files.

日本語の概要は準備中です。原文の説明を表示しています。

Linux privilege escalation playbook. Use when you have low-privilege shell access and need to escalate to root via SUID/SGID binaries, capabilities, cron abuse, kernel exploits, misconfigurations, or credential harvesting on Linux systems.

日本語の概要は準備中です。原文の説明を表示しています。

Escape from Docker containers to the host system using container misconfigurations, mounted sockets, privileged mode, capabilities abuse, and kernel exploits. Use this skill when testing containerized environments for breakout vulnerabilities during penetration tests. Covers Docker socket mounting, cgroup escapes, nsenter techniques, and Kubernetes pod escapes.

日本語の概要は準備中です。原文の説明を表示しています。

Build persistent multi-agent operating systems on Claude Code. Covers kernel architecture, specialist agents, slash commands, file-based memory, scheduled automation, and state management without external databases.

日本語の概要は準備中です。原文の説明を表示しています。

Drive Chromium from standard Playwright APIs with a real-device fingerprint applied in the kernel, one persistent isolated profile per identity, and a per-profile proxy whose exit IP sets timezone and WebRTC - JavaScript (npm 'anti-detect-browser') or Python (PyPI 'antibrow'). Use when sessions must stay logged in across runs and stay separate, when a scraper or agent is blocked by an incoherent headless fingerprint, when checking ads or pricing from another region, when a page must be reached as a phone rather than a desktop, when automation mints a profile per task, when running several of your own accounts from one machine, or when testing how your own bot detection scores a real device. Also for 'antibrow', 'fingerprint browser', 'multi-account browser', '防关联', '多账号', '安卓模拟', 'Android profile', 'mobile fingerprint', 'temporary profile', 'CreepJS', 'residential proxy', 'browser-use', 'crawl4ai'. MCP control is browser-mcp-agent; isolation is multi-account-isolation.

日本語の概要は準備中です。原文の説明を表示しています。

Give an AI agent its own real browser over MCP tool calls - launch, navigate, click, fill, screenshot, extract text, run JS - with a kernel-level real-device fingerprint and a persistent profile, so the session stays logged in between runs and pages see one coherent device instead of a headless build. No Playwright or SDK code to write. Use when an agent should operate a site itself, when a computer-use / browser-use setup needs a captured real fingerprint rather than a synthetic one, when agent sessions keep losing their login, or when comparing hosted agent-browser services. Also for 'MCP browser', 'browser MCP server', 'let my agent browse the web', 'agent browser control', 'browser-use MCP', 'computer use browser', 'Browserbase alternative', 'Steel browser alternative', 'headless browser detected'. Node (npx) or Python; Windows x64, macOS Intel + Apple Silicon, Linux x64 / arm64. SDK and REST reference is anti-detect-browser; account isolation is multi-account-isolation.

日本語の概要は準備中です。原文の説明を表示しています。