LLM Mart Basic

Clean a raw transcript before ingestion: punctuate, paragraph, label speakers, fix mistranscribed technical terms, and split long recordings by topic. Use this skill when the user drops an auto-generated transcript, subtitle file, podcast or lecture transcript, meeting recording

Draft from the vault: outline from concept pages, keep citations to the user's own sources, and surface where their material disagrees. Use this skill when the user wants to write an article, post, essay or newsletter based on what they have collected, or asks what they could wri

Use this skill when selecting and reporting verification for Skill changes; triggers include Skill change verification, quality gates, and evidence levels.

Use this skill when designing, running, interpreting, or reporting Agent Skill evaluations, selecting cases or judges, and analyzing trigger, benchmark, or regression evidence; triggers include Skill evaluation and evaluation design.

Use this skill when reviewing the contract completeness of Skills, Prompts, metadata, or QA documentation; triggers include Skill prose review, Prompt review, and contract audit.

Use this skill when auditing or trimming process residue from Skills, Prompts, comments, or docs; triggers include process prose cleanup, review residue, and current-state rewriting.

Use this skill when reviewing a complete Skill package for architecture, scope, triggers, independent installation, bilingual consistency, Eval readiness, and evidence boundaries; triggers include Skill quality review and package review.

Use this skill when you need to review acceptance criteria for ambiguity, missing rules, and verifiability; triggers include acceptance criteria review.

Use this skill when you need to design accessibility testing against WCAG, keyboard navigation, and assistive technology scenarios; triggers include accessibility testing and a11y testing.

Use this skill when you need evidence-bounded failure classification, retry/fallback/escalation, state consistency, user notice, and recovery evidence; triggers include Agent 故障恢复 and Agent failure recovery.

Use this skill when you need evidence-bounded checkpoints, heartbeats, resume, cancellation, duplicate submission, timeouts, and resource lifecycle; triggers include 长运行 Agent and long-running Agent.

Use this skill when you need evidence-bounded loop state, plan/action/observation cycles, stop conditions, budgets, repetition, and trace evidence; triggers include Agent 循环 and Agent loop.

Use this skill when you need evidence-bounded memory write/read/update/delete, retention, contamination, isolation, provenance, and forgetting behavior; triggers include Agent 记忆 and Agent memory.

Use this skill when you need evidence-bounded Agent identity, tool/resource scope, approval, denial, escalation, and side-effect boundaries; triggers include Agent 权限 and Agent permission.

Use this skill when you need to test AI agent tool-call contracts, authorization, failures, and side-effect boundaries; triggers include agent tool testing.

Use this skill when you need to test AI agent goals, state, planning, recovery, and safety boundaries; triggers include ai agent testing.

Use this skill when you need AI-assisted testing workflows such as test data generation, root-cause analysis, and prioritization; triggers include AI-assisted testing and AI for QA.

Use this skill when you need to test an AI-enabled product feature for behavior, safety, and user-impact boundaries; triggers include AI feature testing.

Use this skill when reviewing AI-generated unit, functional, API, or end-to-end tests for false confidence, weak assertions, missing risks, or unsafe test behavior; triggers include AI-generated test review and functional test review.

Use this skill when you need evidence-bounded safety policy, abuse categories, refusal/redirect behavior, privacy, escalation, and Human risk decisions; triggers include AI 安全 and AI safety.

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.

Use MCP Inspector to connect to local or remote servers, inspect capabilities, call tools, read resources, test prompts, and diagnose failures before release.

Build an MCP server in TypeScript with focused tools, validated schemas, local and remote transports, Inspector tests, and production security controls.

{{SKILL_ENTRY:refactor}}

{{SKILL_ENTRY:release}}

Review a diff with the reviewer fleet, funneled to one triaged verdict. Targets the current working diff, a path, or an inbound GitHub PR.

Exploratory spike on a throwaway branch — answer a named question with disposable code. Never merges; exits to a findings note or {{CMD:feature}}.

Autonomous sprint — one interactive spec gate, then plan-to-PR execution with every auto-decision SMARTS-scored and logged. Hard gates remain true stops.

Daily repo hygiene — review the day's repo state, then perform the cleanups under per-action confirmation. Fast-forward only, never destructive without a yes.

Show the project's current state at a glance — stage, open tasks, open questions, overrides since the last checkpoint, current branch. Read-only.

Wire codeArbiter's statusline into ~/.claude/settings.json, or remove it.

The sanctioned task-board mutator — add a queued task, start one (flips to in-progress and stamps the date, minting a dotted ID on pick-up), or mark an in-progress task done. The only blessed write to open-tasks.md.

{{SKILL_ENTRY:security-architecture}}

{{SKILL_ENTRY:tribunal}}

Watch a PR's CI to completion — diagnose on red, notify and offer the merge on green. Never auto-merges.

Copy a file OUT of a running sandbox box to the host — host-initiated egress only (docker cp). The reverse, a host→container bind, is impossible by construction.

Tear down a sandbox box — remove its container and named volume. --keep-volume leaves the volume; with no id, prune reclaims any leaked ca.sandbox=1-labeled object. Cached images are retained.

Run a single command inside a running sandbox box and capture a JSON result — exitCode, separate stdout/stderr, and a truncated flag past the byte cap. The scriptable exec seam.

Open an interactive shell inside a running sandbox box at /work/repo. Read-only root, non-root user, no host-FS access — explore the untrusted code interactively, then exit.

Pull an untrusted repo into an ephemeral, host-FS-isolated Docker container — clone into a named volume, build a dep-cached image, run under structural isolation. Network defaults to offline. Requires Docker and nixpacks.

Vet a new or changed third-party dependency for license, provenance, and supply-chain risk before any install runs.

Inspect ADR health read-only; optionally select one ADR with --adr N.

Record user-decided ADRs or inspect their health read-only. Preserve attribution and acceptance evidence.

PiG (Pi in Go) is a faithful Go port of upstream Pi, the TypeScript codebase behind the Pi coding agent. It is a parity-bound translation, not a rewrite: upstre…

An AI Agent that lives in your pocket. Local-first and privacy focused.

Unofficial skill that teaches coding agents to build with TypeSafe AI's Jev: typed decisions, calibrated confidence, and prior art from 150+ community projects.

Adaptive Test-time Learning and Autonomous Specialization

Prediction-market trading engine — Wang Transform pricing on 291K+ contracts; paper-traded across Kalshi · Polymarket · Solana DFlow (Jito bundles) · 633 tests

Knowledge Management for Humans and Agents

Open-source Claude Cowork / Codex / WorkBuddy alternative — a local-first AI office agent that turns one request into real PPTX, DOCX, XLSX and HTML files. Runs…

DeepAgent Code: AI coding agent with persistent memory and control plane