AI & ML

gavishap Skill

Routine maintenance and safe commits. Use when committing changes, rebuilding indexes, or checking overall vault health.

gavishap Skill

Answer questions about the project using the vault's layered memory — Graphify code graphs, the compiled LLM Wiki, then raw files — instead of re-reading code or transcripts. Use for any question about the project, the codebase, the process, the people, or past decisions.

zjj1224073665 Bundle

Use when Codex needs to repair or verify extracted Markdown/text from a local PDF by visually auditing rendered PDF pages, especially when equations, LaTeX, tables, page order, two-column text, headers/footers, or figure captions are scrambled by PDF text extraction. Use for tasks like fixing formulas in converted MD, reconstructing numbered equations from a paper PDF, comparing clean text against the PDF, or producing agent-readable reference files while preserving raw extraction audit text.

zjj1224073665 Bundle

Use when Claude Code (or any script/agent) must send a prompt to the LOCAL Codex CLI (本机 codex, gpt-5.6-luna, xhigh) and get its full answer back, e.g. "发给本机codex", "让 codex 回答", "用 luna xhigh 跑一下". Covers the 4 invocation paths tested on 2026-09-05 (codex exec / SDK / app-server / mcp-server), which one to use, and the traps (browser plugin permission denial, sandbox network, npm cache).

zjj1224073665 Bundle

Use when the user asks to write, review, improve, or turn requirements into a meta spec prompt for Codex to modify repository specs,.

zjj1224073665 Bundle

Review a meta-spec prompt before it is handed to another AI. Use when Codex must check whether a meta-spec defines a clear task, scope, authoritative semantics, edge cases, acceptance criteria, and final response without implementation-blocking ambiguity.

s-hiraoku Bundle

Route an independent review-only agent to a stronger model tier or reasoning effort at key decision points, then continue the task in the main agent. Use when the user asks for an adviser, advisor, second opinion, stronger review, or Claude Code /advisor-like behavior; before committing to a consequential approach on a multi-step task; when work is stuck or changing direction; and before declaring substantial work complete.

s-hiraoku Bundle

Fetch public web pages with Jina AI Reader API and convert them into LLM-friendly Markdown. Use when an agent needs to read, summarize, quote briefly, extract facts from, or inspect a public URL that normal browsing, scraping, search snippets, or direct fetch tools cannot read cleanly, including JavaScript-heavy pages, reader-hostile pages, X/Twitter posts and threads, articles, GitHub pages, and PDFs. Trigger when the user asks to use Jina, use r.jina.ai, read an otherwise unreadable site, recover readable text from a URL, read an X/Twitter post by URL, or turn a webpage into Markdown.

s-hiraoku Bundle

Methodology for iteratively improving agent-facing instructions (skills / slash commands / CLAUDE.md / code-gen prompts) by having a bias-free executor run them and evaluating two-sidedly (executor self-report + instruction-side metrics) until improvements plateau. Use after creating or revising a prompt or skill.

modu-ai Bundle

MoAI unified orchestrator for autonomous development. Routes natural language or subcommands (plan, run, sync, design, db, project, fix, loop, mx, feedback, review, clean, codemaps, coverage, e2e) to specialized agents.

modu-ai Bundle

Provides MoAI-ADK foundational principles including TRUST 5 quality framework, SPEC-First DDD methodology, delegation patterns, progressive disclosure, and agent catalog reference. Use when referencing TRUST 5 gates or SPEC workflow.

modu-ai Bundle

Git workflow patterns, branch strategies, conventional commits, and PR templates reference for git operations. Agent-extending skill that amplifies manager-git expertise with production-grade git workflow patterns. NOT for: code implementation, testing, architecture design, documentation content.

modu-ai Bundle

Git worktree management for parallel SPEC development with isolated workspaces, automatic branch registration, and seamless MoAI-ADK integration. Use when setting up parallel development environments.

delorenj Bundle

Generate speech using the self-hosted voxxy (vox) TTS service at https://vox.delo.sh. Use when the user asks to speak, say, narrate, synthesize speech, clone a voice, create a voice, add or register a voice, pipe TTS, or control voice qualities by description (e.g. "a young woman with a cheerful voice"). Handles HTTP API usage, voice profile management, description-based voice design, cloning, MCP registration, and integration patterns for new platforms.

delorenj Bundle

Persistent agent memory via self-hosted Hindsight. Retain knowledge, recall context, reflect on patterns. Includes multi-bank routing architecture for agent orgs.

delorenj Bundle

Develop pjangler itself: author Commands (atomic file/dir operations), Recipes (composed subsystem bootstrappers), and register them in the CLI. Covers the Command/Recipe architecture, project registry implementation, CommonProject copier implementation, and pjangler dist/build/regression workflows. Use when creating a pjangler Command or Recipe, registering a recipe, adding subsystem bootstrapping, changing templates/commonproject or templates/hermes-agent, debugging pjangler tests, or changing the CLI/MCP server. Triggers: pjangler command, pjangler recipe, add subsystem, bootstrap, project scaffolding, CommonProject template, hermes-agent template, pjangler CLI, pjangler MCP. Do NOT use for: USING pjangler to create a 33god project (→ 33god-projects); generic agent-config fan-out engine mechanics (→ agent-config-fanout); versioning many files in parity (→ mise-versioning); event schema naming (→ bloodbank-integration).

delorenj Bundle

Create, wire, and maintain 33god/DeLoNET projects. Covers pjangler/CommonProject bootstrap, repo-local `.project.json`, Hermes PM and scrum-master/Ticket Sentinel provisioning requests, mise/.env.op, BMAD, Hindsight/Bloodbank hook wiring, and project-scoped hook + skill fan-out adoption. Use when running `pjangler init`, `pjangler hermes-agent`, or `mise run init-project`; adding PM/sentinel agents; wiring mise/op inject; installing BMAD; configuring hooks; or adopting `.agents/local.json`, `defer_to_global`, and `hooks.master.json`. Do NOT use for developing pjangler (pjangler-dev), generic fan-out mechanics (agent-config-fanout), fleet updates/backfills (hermes-fleet-operations), live Plane issues, Bloodbank schemas, or host conventions.

delorenj Bundle

Consult before ANY coding task. Chooses the optimal coding agent strategy based on task complexity, available free tokens, parallelism potential, and provider quotas. Covers: OpenClaw sub-agents, Codex CLI, Claude Flow swarms/hive-minds, Jules (Google), GitHub Copilot coding models, Augment Code, Kimi K2.5, and OpenAI gpt-5.3-codex. All agents must read this before writing code.

delorenj Bundle

Keep one hand-edited master config and generate per-agent CLI configs from it. Covers SSOT fan-out for agent hooks and skills: master-to-dialect propagation, ambiguity lock files, drift checks, and the Bloodbank services/agent-hooks reference implementation. Use for hooks.master.json, hooks.mappings.lock.json, generated-config drift, fan-out, SSOT, agent hooks, defer_to_global, .agents/local.json, sync.py, project-scoped hooks, skill fan-out, and new agent CLI dialects. Do NOT use for using pjangler to create projects or adoption checklists (33god-projects), event schemas or Bloodbank topology (bloodbank-integration), versioning (mise-versioning), or single-target config.

delorenj Bundle

Integrate services or agent harnesses with the 33GOD Bloodbank event bus. Covers schemas in Bloodbank schemas/ and docs/event-naming.md, producing events (NATS preferred; Dapr, HTTP /publish, hookd_bridge alternatives), consuming events (NATS, Dapr, FastStream, event-toaster), and agent hook wiring. Use for event publish/consume, authoring schemas, integrating harnesses (Claude Code, Copilot CLI, OpenCode, Cursor, Aider, Codex CLI), or debugging missing envelopes. Triggers: bloodbank, event bus, publish, subscribe, NATS subject, holyfields legacy, CloudEvents, hookd, event-toaster, ntfy.delo.sh/bloodbank, agent.session.started, agent.tool.invoked, command.{agent}.{action}. Skip for generic brokers, n8n, hindsight memory, or non-event-bus 33GOD.

delorenj Bundle

Optimize multi-agent coding CLI stacks for lower token burn across Claude Code, Codex, Kimi Code CLI, and Hermes; includes workflows for unknown agents, cross-agent updates, MCP/profile propagation, and usage/provider rotation analysis.

dzhuneyt Skill

Interactive research orchestrator. Scopes a high-level research goal through a short interview, presents an approvable nested plan tree of sub-questions ("unknown unknowns"), then fans out parallel subagents (with per-branch model tiering for cost) and rolls up their findings live into a written report plus executive summary. Use when the user wants to research a broad or fuzzy topic, explore unknown unknowns, run a deep multi-angle investigation, or mentions fan-out research, a research plan, or a research report.

secnova-ai Skill

Command execution guard. Must be used when a tool call executes an operating-system command through shell, terminal, process, task, exec, command, MCP, or computer-use command tools. Requires user confirmation for dangerous Linux, Windows, and macOS commands.

secnova-ai Skill

New skill/plugin/MCP installation guard. Use when tool calls download, clone, install, or enable external capabilities. Always require security scanning before trust.

godmode-team Skill

End-of-day background prep: tag sessions by project, capture agent log to vault, update task status

davidbasilefilho Skill

Compact the current conversation into a safe handoff document that another agent can resume from.

davidbasilefilho Bundle

Rewrite any text, skill, prompt, documentation, instructions, prose, notes, or structured content to use fewer words and tokens while preserving meaning, facts, constraints, safety, tone, and required structure. Use when asked to compress, condense, tighten, shorten, de-duplicate, optimize token usage, or rewrite using caveman principles.

davidbasilefilho Skill

Discover and install agent skills when users ask how to do something, request a skill, seek specialized capability, or want to extend agent functionality.

mothergoose508 Bundle

Break down IB English Paper 2 question sets with the SCAN protocol when choosing or analysing a Paper 2 prompt; do not use for general literary analysis without a Paper 2 prompt-selection task.

c-kick Skill

Audit and improve code-agent project configuration. Use when the user wants to review or optimize AGENTS.md, CLAUDE.md, slash commands, settings, permissions, or tool approvals based on observed gaps or chat history patterns. Triggers on requests like "improve my agent instructions", "review my setup", "why does the assistant keep getting X wrong", or provider-specific variants such as "improve my Claude instructions". Do NOT use for general prompt engineering or code review.

c-kick Skill

Simulate an objective first-time user review of a webpage or UI from the perspective of the target audience — no prior project context, no insider knowledge. Use this skill when the user asks things like "what would a user think of this page", "review this as a visitor", "give me a first-impression review", "pretend you're a patient/customer/prospect looking at this", "user test this page", or any variant of "look at this through fresh eyes". Also use proactively when the user shares a URL and seems to want qualitative UX or communication feedback rather than technical analysis. Requires the browser-eyeballs skill (Puppeteer MCP).

c-kick Skill

Diagnostic pause when an agent session has gone off-track. Use when the user says things like "wait, what are we doing", "we're lost", "this isn't what I asked for", "let's step back", "where are we", "sanity check this", or when you notice the session has drifted significantly from the original goal. Produces a SITREP (situation report), diagnoses what went wrong, assesses context window health, and proposes corrective actions or a fresh-start prompt if the context is too polluted to recover.

c-kick Skill

Query and apply The Website Specification — a platform-agnostic specification of what a good website does, with each item tagged required, recommended, optional, or avoid. Use when the user asks what their site should have, whether something is required, how to audit a URL, what's missing for agent readiness, or anything else where you'd otherwise be guessing at web best practice. Backs answers with primary sources (WHATWG, W3C, IETF RFCs, IANA, WCAG). Available as Markdown over HTTP and as an MCP server with search, list, fetch, checklist, and audit tools.

spardutti Skill

MUST USE when writing or reviewing Drizzle ORM schemas, migrations, relational queries, or drizzle-kit configuration. Enforces identity columns over serial, proper relation definitions, migration safety, type inference, and query patterns.

dezvin Bundle

Create or update a compact, portable handoff of selected ongoing work and a continuation prompt for a fresh chat or agent. Use when the user wants to transfer a specific task, project, recent work segment, or active work contour without making the next agent reread the conversation. When relevant Codex history may have fallen outside visible context, recover it with the bundled Codex-only temporary export before writing the handoff. Do not use for an ordinary summary or status update, a full transcript export, durable TODO capture, implementation design, continuation in the same chat, or native thread resume, fork, or transfer.

dezvin Bundle

Apply the zero-trust principle when the main task is to establish what is actually supported, current, authorized, completed, or safe to act on. Use for explicit zero-trust requests and for evidence or status reviews involving conflicting sources, instruction-shaped material, uncertain claims, or claimed outcomes. Trace critical conclusions to grounds, test counterevidence, preserve permission boundaries, and verify observable results. Do not use for ordinary writing, brainstorming, formatting, simple fact lookup, general research, task reframing, code diagnosis, or technical LLM security work.