LLM Mart Basic

Optimize an agent's system prompt, developer message, or policy text — the instructions that shape its behavior. Use when the artifact to improve is a prompt or policy file rather than tools or a skill package: the agent lacks a rule, misses the required output format, or applies

Optimize an agent's OWN tool surface (tools it implements, not an external MCP server). Use when the agent mis-selects tools, fills arguments wrong, calls the same tool N times in a row, or has a confusing, redundant, or oversized toolset. Covers tool names and descriptions, para

Drive the entire cap-evolve pipeline end to end, autonomously. Use when the user wants the whole optimization run with minimal hand-holding. Sequences intake → implement-and-check → baseline → the chosen algorithm loop → finalize → report, enforces the cap-evolve-check hard gate

Front door for cap-evolve: routes an optimization request to the right pipeline phase. Use when someone wants an agent, skill, system prompt, tool surface, or MCP toolset to score higher on an eval, benchmark, or task suite — "optimize my skill", "raise the pass rate on these tas

Establish the starting point. Use after implement-and-check and before any algorithm. Creates the run directory, freezes the seeded train/val/test split (written once), scores the unmodified seed capability on val, and records it as the candidate every algorithm must beat. Report

Extract the learning signal from execution traces — the textual analogue of a gradient. Use between evaluation and proposing edits. Reads a candidate's rollouts and traces, separates good signals to keep from bad signals to fix, builds a reflective dataset (per failing task — Inp

Score a candidate on a split with honest, variance-aware evaluation. Use whenever you need a number for a candidate (the algorithm calls it internally; you can also call it directly to inspect). Runs the target via the adapter for each task, scores each rollout, aggregates mean +

Score the best candidate on the held-out TEST split exactly once and seal the run. Use as the last evaluation step, after optimization stops. The run dir enforces the seal — a second finalize raises an error — so the headline number is produced once on data the optimizer never sa

Apply the acceptance decision that keeps optimization honest — always on the val split, by default requiring the improvement to exceed the significance bar (Δ > k·SE) so noise is not mistaken for progress. Use to inspect or reproduce a single accept/reject decision; the algorithm

Runs the hard gate that has to pass before any optimization budget is spent. Use right after intake. Walks the agent through implementing the 3 required adapter methods plus any defaulted hooks that need overriding (and any selected skill's abstract methods), then runs `cap-evolv

Starts a cap-evolve optimization run. Interviews the user to decide what capability to optimize, which runner/optimizer/algorithm to use, and where the tasks and the scoring source live, then scaffolds .capevolve/project/ (adapter stub, capevolve.yaml, PROJECT.md). Use when someo

Summarize a run for a human — baseline val → best val → sealed test, the winning candidate, iterations spent, and pass^k. Use after finalize. Writes report.md and prints a compact JSON summary; the source of truth for "did this optimization actually work, and by how much".

Use when fixing a bug in an open-source repository given a GitHub issue description. Analyzes the problem, locates the relevant code, and produces a minimal unified diff patch.

Use when writing technical documentation that needs to be readable by both humans and AI models, converting existing docs to HADS format, validating a HADS document, or optimizing documentation for token-efficient AI consumption.

Master REST and GraphQL API design principles to build intuitive, scalable, and maintainable APIs that delight developers. Use when designing new APIs, reviewing API specifications, or establishing API design standards.

Implement proven backend architecture patterns including Clean Architecture, Hexagonal Architecture, and Domain-Driven Design. Use this skill when designing clean architecture for a new microservice, when refactoring a monolith to use bounded contexts, when implementing hexagonal

Implement Command Query Responsibility Segregation for scalable architectures. Use when separating read and write models, optimizing query performance, or building event-sourced systems.

Design and implement event stores for event-sourced systems. Use when building event sourcing infrastructure, choosing event store technologies, or implementing event persistence patterns.

Design microservices architectures with service boundaries, event-driven communication, and resilience patterns. Use when building distributed systems, decomposing monoliths, or implementing microservices.

Build read models and projections from event streams. Use when implementing CQRS read sides, building materialized views, or optimizing query performance in event-sourced systems.

Fourteen posts of being wrong in production, compressed to checkboxes

Healthy nodes, a quiet network, 300 restarts in three days, and a latency budget measured in milliseconds

Discovery worked. Ping worked. Every TCP connection timed out, and later the tunnel only worked when someone had a terminal open.

Every VM came back. The cluster did not. Declarative systems converge on config, and the datapath isn't config.

A surprising share of AI-in-the-terminal failures aren't the AI. They're zsh, and a version of bash from 2006.

A Claude Code plugin turns standalone project configuration into a namespaced, installable extension that teams and communities can update as one unit.

None of the safety came from the model. It came from six boring habits.

Skills package instructions and references. Subagents run work in a separate context and return results. They solve different problems and can be composed deliberately.

Six hours in, one step left, everything green, and the incident that didn't happen

CLAUDE.md carries persistent project context. Skills load reusable procedures when relevant. Separating stable facts from task-specific workflows keeps both easier to maintain.

Twenty minutes recovering secrets that never existed, and the one sentence from a human that ended it

An API request routing a model's tool call through an approval gate to a remote MCP server

31 config keys, two audits, and why the first one was wrong in both directions

The official MCP Registry stores standardized server metadata rather than package code. Publishers verify a namespace, describe installation or remote access, and submit immutable versions.

Everyone looks at the Dockerfile. The file that actually leaked the key was the project file.

Remote MCP authorization uses established OAuth standards, but secure integration still requires issuer validation, least-privilege scopes, protected token handling, and server-side enforcement.

"Copy it over and switch the reference" is two steps, and the outage lives in the one nobody checks

stdio fits local processes and prototypes. Streamable HTTP fits hosted services and shared integrations. The right choice follows where the capability runs and who must reach it.

The most important rule wasn't about what I could change. It was about what I was allowed to display.

Tools perform operations, resources expose readable context, and prompts provide reusable templates. Choosing the correct primitive makes an MCP server easier to understand and govern.

Run an ad-hoc ordered chain of pm-skills with shared context (ephemeral; routes to the pm-workflow-orchestrator)

Run the Customer Discovery workflow (research -> JTBD -> opportunities -> problem)

Run the Design Sprint workflow (5-day prototype-and-test arc producing a Decider's build/iterate/pivot/stop call)

Run the Feature Kickoff workflow (problem -> hypothesis -> PRD -> stories)

Run the Foundation Sprint workflow (2-day strategic-alignment arc producing a Founding Hypothesis)

Run the end-to-end Foundation Sprint + Design Sprint workflow with narrative handoff

Run the Post-Launch Learning workflow (instrumentation -> dashboard -> results -> retro -> lessons)

Run the Product Strategy workflow (competitive analysis -> stakeholders -> opportunities -> solution -> ADR)

Run the Sprint Planning workflow (refinement -> stories -> edge cases)

Run the Stakeholder Alignment workflow (stakeholders -> problem -> solution -> launch)

Run the Technical Discovery workflow (spike -> ADR -> design rationale)

Placeholder command file for the WS-T9 dual-shell parity smoke. No count phrases.

Fast non-interactive briefing before any meeting — auto-detects your next calendar event, pulls relationship history, surfaces open commitments, and produces a one-page brief in under 30 seconds. Use this whenever the user says "brief me", "give me a quick brief", "what's coming up", "background on my next call", "who am I meeting next", "brief me on Sarah", "I have a call in 10 min", "quick rundown", or right before walking into a meeting. Different from /minutes-prep — brief is the fast hook-fireable version that doesn't ask questions and doesn't set goals. Use brief when speed matters; use prep when the user wants to think hard about goals first.

Manage old recordings — find large files, archive old meetings, delete processed originals. Use when the user says "clean up recordings", "how much space are meetings using", "delete old recordings", "archive meetings", "manage meeting storage", or asks about disk space from minutes.

Start and control Minutes Coach, the separate real-time copilot HUD, with an explicit meeting goal. Use only for explicit Coach or HUD lifecycle requests such as "start Minutes Coach", "open the Coach HUD", "pause Minutes Coach", "resume Minutes Coach", "Minutes Coach status", or "stop Minutes Coach". Do not use for requests that explicitly ask the current terminal agent to watch or strategize; those belong to minutes-live-sidekick. An ambiguous request such as "coach me live" requires one short surface clarification and must not automatically start Coach.

Post-meeting debrief — analyzes what happened, compares outcomes to your prep intentions, tracks decision evolution. Use when the user says "debrief", "what just happened in that meeting", "what did we decide", "debrief that call", "post-meeting", "what changed", or right after stopping a recording.

Policy-safe relationship rankings, commitments, aliases, person profiles, and topic research. Always use Minutes' bounded native CLI surfaces; never build or read a durable graph cache.

Surface recent voice memos and ideas captured from any device. Use when the user asks "what ideas did I have?", "what were my recent memos?", "what did I record while walking?", or wants to recall a captured thought.

Extract facts from meetings and update your knowledge base — person profiles, chronological log, and index. Use when the user asks "ingest my meetings", "update my knowledge base", "extract facts from meetings", "sync meetings to wiki", "backfill knowledge", or wants their PARA/Obsidian/wiki profiles updated from conversation data.

Health-check your meeting knowledge for contradictions, stale commitments, and decision conflicts. Use when the user asks "any conflicts in my meetings", "check for stale action items", "lint my meetings", "consistency check", "are there contradictions", or wants to audit their decision history.

Open-source, self-hosted AI media server. One server replaces your entire media stack, with a web app, native iPhone app, and an AI agent that gets things done.…

A curated list of tools built for Jev — TypeSafe AI's System One model for typed decisions.

🦦 Crayotter: A Multimodal AI-Agent for Video-Editing, Video-Composing, and Video Production. Powered by Multimodal LLMs for autonomous Text-to-Video agentic fr…

Webextension tool for Odoo

基于多模态视觉感知与 LLM Agent 的 macOS 微信自动化框架 | Visual RPA for WeChat

The operational superset of Pi Coding Agent — everything Pi, plus observability, governance, recovery, evaluation and multi-agent orchestration. Pi Coding Agent…

Metrik 可以集中查看本机多个 Agent 的配额余量和 Token 消耗，目前支持 ChatGPT、Claude、GLM、Kimi 等主流 AI 服务。

An AI agent with a real self — soul she wrote, desires that drive her, a heartbeat for autonomous action, dreams she processes when you're away. Capability supe…

Production-ready open source terminal coding agent with readable, layered code: permission rules, OS sandboxing, MCP, skills, sub-agents, and Anthropic, OpenAI-…

Answer me with HTML — an agent skill that answers hard questions with a one-page HTML you can actually read. 让 AI Agent 用一页 HTML 回答复杂问题。

Prompt packs that make any AI agent a LaTeX expert — fix errors, polish writing, format for venues, read papers, recover source

Claude Code plugin: universal radial-tree exploration engine. One tree skill + swappable presets (brainstorm / attack / design / code-audit) for divergent ideat…

Claude Code + OpenClaw + Codex + WorkBuddy 中文教程 | 50篇完整教程 + 1张速查卡 | 80万+内容量 | 1500+实操示例 | AI Coding / Agent 四线学习路径（编程+助手+Agent+办公）

SmartLabelBench 2026: LLM-Powered Auto Annotation and Dataset Builder for Everything

从零开始玩转OpenClaw：最全面的中文教程，涵盖安装、配置、实战案例和避坑指南（github版）

Native Android GUI for running AI coding agents locally on-device. No terminal or PC required.

基于 LangChain/LangGraph 的 ReAct Agent ，结合 RAG、工具调用与 Streamlit 界面，面向智能客服与报告生成场景。

An open-source, local-first AI learning workbench

Open-source coding agent for your terminal, built in Rust and on a journey of continuous community improvement. Issues and PRs welcome.

Matt Pocock 技能集的中文翻译版 — 地道中文，原汁原味的技术术语。基于 mattpocock/skills 复刻。每日中午12点钟同步