Results for “tail-latency”

Tail, read, or search a Stata log file from a previous command or background task.

Guides optimization of Qdrant query latency by tuning segments, memory, quantization, and search parameters.

Trains, evaluates, and runs inference for Mask Auto-Label (MAL) weakly-supervised segmentation models using ViT-MAE backbones with minimal point or box annotations.

Analyzes Perfetto traces to find the root cause of latency, memory, or jank issues in Android apps.

Reduce per-token latency on Jetson vLLM servers by appending speculative decoding configuration, with guidance on when to enable and how to benchmark the improvement.

Anchors work at the current moment and places events as spatial distances ahead or behind to improve temporal reasoning.

Applies four disciplined mindsets—audit, debt, help, review—to cut complexity, track deferrals, surface reference, and catch over-engineering in codebases.

Tracks crypto and AI narratives with mindshare scores, velocity arrows, phase labels, named drivers, and position calls. Detects phase transitions and reflexivity signals.

Patent prior-art and landscape intelligence skill — not generic patent help. Commits to one of five sub-use-cases via forcing intake (novelty search / freedom-to-operate / competitive landscape / acquisition diligence / litigation prior-art) before any search runs. Searches Google Patents, Espacenet, USPTO, and optionally Lens.org for citation-graph signals. Output is an editable Word document (.docx) with verdict, ranked closest art (claim-text extracted), CPC-class-aware landscape, family-resolved hits, geographic coverage, FTO flags where applicable, strategy recommendations, and full audit log. Use when the user asks for patent searching or analysis (e.g., 'prior art search for [invention]', 'freedom to operate analysis for [product]'). Produces search signal, not legal advice — always recommends consulting a patent attorney before filing or licensing decisions. Trademark, copyright, and trade-secret questions are out of scope.

Shares memory between agents at the representation level by compacting the orchestrator's KV cache for efficient worker handoff, reducing token costs without summarization or retrieval.

The layers trade timeliness against definitiveness.

Layer alternative channels / scrapers / data sources with explicit fall-through, where each rung covers a different failure mode. Use when interacting with the world (sending a notification, fetching from a third party, detecting a state change, authenticating) and the single canonical path could fail in known ways. Distinct from retrying — retries handle transient failures; fallback chains handle modal ones (channel down, API gone, cookie expired). Three layers tops; the terminal rung is loud.

YAGNI-first code-minimalism ladder — before writing code, check in order: does this need to exist? already in the codebase? stdlib? native platform feature? installed dependency? one line? Cuts unnecessary code/tokens while never trading away validation, error handling, security, or accessibility. Triggers on: 'ponytail mode', 'write minimal code', 'don't over-engineer this', 'yagni', 'is there already something for this', 'use native instead of a library', 'keep this simple', 'fewest lines possible', 'lazy senior dev mode'.

Runs a structured three-question lazy-check on significant output before submission, forcing root-cause analysis, alternative consideration, and outcome verification.

Official GSAP skill for timelines — gsap.timeline(), position parameter, nesting, playback. Use when sequencing animations, choreographing keyframes, or when the user asks about animation sequencing, timelines, or animation order (in GSAP or when recommending a library that supports timelines).

Use when planning a POPL campaign calendar — backward-scheduling theory and mechanization from the July deadline, riding the October notification into the conditional-acceptance revision and artifact evaluation, landing the January conference, and retargeting across the SIGPLAN/PACMPL deadline wheel if July slips.

Caching strategies across the RAG stack. Semantic caching with GPTCache and LangChain, Redis-based embedding-similarity cache, cache key design, TTL/invalidation, partial caching (cache retrieval only), provider-native prompt caching (Anthropic, OpenAI), and hierarchical L1/L2 caches. USE WHEN: user mentions "semantic cache", "GPTCache", "LLM cache", "prompt caching", "Redis vector cache", "cache invalidation for RAG", "reduce LLM cost", "latency reduction LLM" DO NOT USE FOR: retrieval accuracy - use `rag-patterns`; groundedness checks - use `rag-guardrails`; incremental indexing - use `rag-production`

Track elapsed time from a set epoch with tamper-evident locking, supporting count-up, count-down, and count-between modes, milestones, and email notifications.

Use to design VIP or tiered support programmes and to test whether higher tiers actually deliver better outcomes. Trigger for "VIP support programme", "service tiering", queue priority for premium customers, entitlement vs priority, tier inflation, or "does our enterprise tier get faster support".

Design privacy-aware observability for AI agents using traces, spans, structured events, metrics, cost attribution, dashboards, alerts, and investigation workflows. Use when instrumenting an agent, debugging intermittent tool or model failures, defining service-level objectives, analyzing latency or spend, auditing agent decisions, or preparing production monitoring.

Engineering retrospective — generates a data-driven report of development velocity, commit patterns, hotspots, and team insights from git history. Saves results for week-over-week comparison.

Multi-tiered memory management (HOT/WARM/COLD) for context compaction. Invoke ONLY for explicit compaction events: post-`/compact` cleanup, MEMORY.md tier promotion, archive batch, or "trim my context". NOT for general recall (use deep-recall) or routine memory writes (use storage-router). Triggers: "compact memory", "promote to durable", "archive old context", "tier this".

Push the LLM to reconsider, refine, and improve its recent output. Use when user asks for deeper critique or mentions a known deeper critique method, e.g. socratic, first principles, pre-mortem, red team.

TIGER v2 — Multi-scanner trading system for Hyperliquid perps via Senpi MCP. 5 signal patterns (BB compression breakout, BTC correlation lag, momentum breakout, mean reversion, funding rate arb), DSL v4 trailing stops, goal-based aggression engine, and risk guardrails. Configurable profit target over deadline. 12-cron architecture (10 TIGER + prescreener + ROAR meta-optimizer). Pure Python analysis. Requires Senpi MCP, python3, mcporter CLI, and OpenClaw cron system.

Dart static analysis, linting, formatting, and code-generation standards. Use only for analysis_options.yaml, build_runner, dart format, DCM, lefthook, or analyze/format CI failures; defer Dart language, null-safety, Flutter tests, and generic CI questions.

`task-agent`/`review-agent`: use when broker delivery, ordering, acknowledgement, DLQ, backpressure, or replay changes; skip synchronous retry without message semantics.

通用时间抽象接口。当定义时间维度、时序关系、时间演化时调用此技能。

Watches conversations continuously and compresses them into prioritized notes, consolidating and recovering missed sessions with multiple redundancy layers.

Use to analyse touch time vs elapsed time across the resolution path — queue wait, customer wait, back-office wait, and where elapsed time hides real delay. Trigger for "flow efficiency", "touch time vs elapsed time", "where is the time going", "customers waiting on us", back-office bottlenecks, idle time in cases, or end-to-end time much longer than handle time.

当个人或团队因拖延、分心或动力不足，导致重要目标长期停滞不前，需要一种强制机制来激发高度专注和行动力时

Active Directory and network lateral movement assessment. Pass-the-hash, pass-the-ticket, Kerberoasting, AS-REP roasting, NTLM relay, SMB relay, WMI/WinRM/PSRemoting abuse, constrained/unconstrained delegation, RBCD, cross-trust exploitation, and pivoting. Uses impacket, netexec, enum4linux-ng, smbmap, smbclient, bloodhound-python, Responder, ntlmrelayx, mitm6, and ldap-utils. Chains from /post-exploit or /credential-audit.

Use for writing, reviewing, debugging, or testing bounded retry policies in Python with Tenacity, including retry predicates, stop and wait strategies, jitter, exception propagation, callbacks, Retrying, and AsyncRetrying. Do not use for generic loops, scheduled jobs, domain polling without a retryable operation, or operations whose side effects are not safe to repeat.

Turns Lark collaboration noise into prioritized action by triaging chat, approvals, meetings, docs, spreadsheets, calendars, and email with a default read-only counselor mode.

Wire ETL, ingestion, cron, edge-function, and queue jobs correctly. Use for "build a pipeline", "sync X into Y", "nightly aggregation", "cron double-counts", "dedupe", "backfill", "the numbers are wrong after a retry". Bakes in idempotency, atomic writes, data contracts, dead-letter, and observability.

Provides slash commands to start, stop, and message Claude Code relay sessions from chat channels like Telegram, with auto-forwarding and inline button menus.