david-skills

How to make an AI coding agent run on a schedule, loop, or interval — cron jobs, heartbeats, self-starting/self-terminating loops, recurring autonomous checks. Use when the user asks to "run every N minutes/hours", "check something on a regular interval", "schedule a task", "run on a loop", "self-start", "heartbeat", "cron job for an agent", or wants an agent to fire periodically without manual launching. Covers two camps — one-shot agents you wrap in an external clock (Claude Code, Codex, Pi) vs agents with a built-in scheduler daemon (Hermes).

Use only when David explicitly says "track my behavior in this conversation" or clearly asks to log, track, or analyze his behavior in the current conversation. This skill logs each of David's messages to a local behavior database, records decisions with follow-up questions, updates local notes when useful, and surfaces patterns when asked.

Direct browser control via CDP. Use when the user wants to automate, scrape, test, or interact with web pages. Connects to the user's already-running Chrome.

MUST be read ANY time you interact with cmux in ANY way — listing/inspecting/creating/closing cmux workspaces, panes, or surfaces; reading or capturing pane/screen output; sending input or keys to a pane/surface; delegating to, polling, or checking on other agents running in cmux panes/surfaces; building or rearranging terminal layout; cmux browser automation; sending notifications/flashes/status/progress to the sidebar; editing cmux settings; or integrating an agent with cmux hooks. If your command starts with `cmux ` or touches a cmux workspace/pane/surface/agent, read this FIRST. Triggers on "cmux", "in this workspace", "this pane", "the other agent", "delegate to", "check on the agent", "send to the pane". macOS only (14.0+).

Explain and write effective instructions for OpenAI Codex's `/goal` feature — the persistent self-checking agent loop (plan → act → test → review → iterate). Use when the user mentions Codex `/goal`, "goal loop", wants to kick off a long-running autonomous Codex run, asks how to write a goal prompt, or wants a one-paragraph goal instruction drafted.

Use when David is preparing for or running a consulting call (lawyer, accountant, tax advisor, etc.) and wants to manage a question list — adding, marking as asked, moving between Main/Potential, or tracking answers. Triggers on consulting call, advisor call, prep questions, mark as asked, add question to call.

Use when David has raw material (podcast transcript, interview, research notes, finished recording) and needs to find the angle/framing for a YouTube video. Funnels from many angles → one-sentence framings → chosen direction before any outline work begins.

Read-only exposure audit of the user's machine and projects for a CVE, breach, malicious package, or other security advisory, then write a structured report to a local audit folder. Use when the user shares a breach/CVE/malware/supply-chain advisory and asks if they're affected, says "scan my system for X", "are we affected by Y", "check if I'm vulnerable to Z", or requests any hack/breach/cyber/vulnerability audit.

Use when David wants to score or rate the importance of his Todoist tasks in docs/data/todoist-task-dimensions.db. Presents tasks 7 at a time (A-G), David gives scores (0-10), AI saves them. Also updates brain/ and docs/data/user-behavior.db.

Use when David wants to plan his day, create a priority list, figure out what to work on, or asks "what should I do today." Triggers on day planning, priority list, what to work on, plan my day, what's most important.

MUST be read ANY time another AI agent is involved — sending a prompt to or polling a Pi/Codex/Claude Code/Hermes/OpenCode agent, driving an agent running in another cmux pane/surface/terminal/tmux window, delegating or relaying ANYTHING to an agent in the same cmux workspace, spawning a sub-agent, or orchestrating agent-to-agent work. Read it BEFORE the first message you send to another agent. Covers the critical rule that prompts sent to a TUI agent must be ONE single line (newlines = Enter = fragmented/cut-off prompts), short polling cadence, which agent to pick, and per-agent characteristics. If you are about to type `cmux send`/`tmux send-keys` to any agent, this skill applies.

Distribute a newly-created agent skill across all 4 canonical skill folders on a local machine so Claude Code, Codex/OpenAI Agents, Pi Agent, and Hermes all see it. Use whenever the user says "give this skill to all agents", "distribute this skill", "sync skills across agents", or after creating a new skill that should be globally available. Covers the .claude/skills symlink and the ~/.pi/agent/skills (NOT ~/.pi/skills) trap.

How to write effective agent skills — what to do, what not to do, anatomy, progressive disclosure, design patterns, anti-patterns, testing, security. Read this whenever a skill (Claude Skill, Agent Skill, SKILL.md) is being created, edited, reviewed, or debugged. Use when the user says "create a skill", "new skill", "update this skill", "improve a skill", "why isn't my skill triggering", or anything else involving authoring or editing SKILL.md files.

Use when David wants to mass-migrate tasks out of Todoist into brain/good-ideas.md. Filters all tasks against David's current priorities, keeps only actionable tasks, moves everything else. Triggers on purge, migrate, clean todoist, too many tasks, overwhelmed, move to endless ideas.

Use when David wants to plan tomorrow's 6 priorities the evening before. The Ivy Lee Method — pick 6 things, rank by importance, execute in order starting with

Create a specialized CLAUDE.md (+ AGENTS.md symlink) inside a specific folder to give future agents folder-scoped context. Use when David asks to create a CLAUDE.md for a folder, write folder instructions, or add agent context to a directory.

Improve the visual quality, cohesion, and finish of an existing frontend without changing the product direction. Use when the user asks to polish, refine, tighten, elevate, clean up, or make a UI feel more premium, clearer, or more consistent across pages, components, spacing, typography, color, motion, states, and responsive behavior.

Use when David is interviewing a candidate for an office hire. Parses the latest candidate application CSV from ~/Downloads, builds a TLDR + flags, copies the master question list into a per-candidate file, generates 5–6 sharp custom questions, tracks asked questions + live notes during the call, and runs a post-interview debrief. Triggers on interview prep, parse the candidate, candidate TLDR, build interview questions, custom questions for this guy, mark question as asked, reset to un-asked, next candidate, compare candidates, post-interview verdict, debrief.

Read all 10 main context files (docs/main-files.md index), then ask David 1-2 broad discovery questions per file about the DOMAIN (not the file contents) to surface new reality. Update each file based on his answers. Use when David wants to sync/refresh his core context.

How to reliably open a markdown file in a cmux right pane without it rendering blank. Use whenever opening/showing/rendering a .md file in cmux's right pane via `cmux markdown open`. Covers the move-surface blank-render bug and the only two reliable approaches — open it correctly the first time, or close the existing right pane(s) first then open a fresh right pane.

Design and outline New Society classroom modules. Use when David asks to work on NS course outlines, module layouts, classroom structure, or Section 2/3 design.

Pre-production for New Society course modules using tldraw. Use when David wants to design, outline, or build visual canvases for NS modules before recording. Triggers on NS module prep, module visual, module canvas, pre-production, recording prep.

Use when David asks where Obsidian is installed, where his notes/vault/journal/writing live, or wants to read/find/edit a note from his Obsidian vault. Triggers on "Obsidian", "my notes", "my vault", "my journal", "private writing", "where do I write".

Use when David wants to clean up, review, or triage his Todoist tasks. Batch review process - 4 tasks at a time, David decides, AI executes.

Use when David wants to write copy or scripts for paid ads (Meta Ads — FB + IG, Google Ads, etc.). Triggers on ad copy, ad script, ad creative, hooks for ads, paid ads, Meta ads, FB ads, Instagram ads.

YouTube video pre-production. Use when David is preparing a video that involves a tool, repo, product, build, or use-case demo. Two phases — (1) VET the tool actually works before committing recording time, then (2) BUILD the simplest reproducible steps David records from. Triggers on pre-production, pre-prod, test before filming, vet this tool, try it out, sandbox test, viability check, does this actually work, use case testing, figure out simplest steps, reproducible steps, demo prep, build the use case.

Use when David is building or reviewing a YouTube video presentation (Google Slides) slide-by-slide. Covers the Firebase Studio alternating explain/show structure, the coverage tracker, and the bullet-by-bullet iteration loop. Triggers on presentation slides, slide outline, video deck, Google Slides, slide structure, "what's the next slide", "give me options for this bullet", presentation review.

Verify a bank's monthly CSV and PDF exports agree before trusting either for bookkeeping. Use when reconciling bank statements, importing monthly bank exports, or building a local ledger from raw bank data.

Use when David is mid-day, feels scattered, and wants help collapsing chaos into focus. Triggers on "reduce chaos", "gain clarity", "help me focus on rest of day", "I'm pulled in too many directions", "what should I actually do now". NOT for fresh morning planning (use day-plan for that).

Use when David wants to build a structured strategic document by answering questions (life priorities, goals docs, framework files, ranked lists, principles, reviews). Interview one question at a time, patch the file after each answer, then re-ask. Not for day planning (use day-plan).

Create Vapi voice assistants and place outbound phone calls via the Vapi MCP — cold outreach, SDR, appointment setting, callbacks.

Write a single-paragraph Deep Research prompt to hand to a human researcher (or a deep-research AI). Use when the user wants a research brief, a "deep research prompt", a one-paragraph task for a researcher, or asks "what should our researcher look for". Produces ONE tight paragraph with full context, numbered sub-questions, and per-finding output format.

Score any AI model on the DeepSWE coding-agent benchmark via the OpenRouter API. Use when the user wants an independent, reproducible coding-agent eval — "run DeepSWE", "benchmark this model on DeepSWE", "score model X on the coding benchmark", "test a model via OpenRouter on DeepSWE", or to verify vendor-reported coding scores. Covers setup, the OpenRouter wiring for mini-swe-agent, single-task / subset / full 113-task runs, and leaderboard submission.

Commit and push agent-skill changes to the user's private skills GitHub repo, rooted at ~/.agents. Use after creating or updating any skill, when David says "push the skill", "push skills to github", "save the skill to my repo", or "update the skills repo". Handles staging, committing, pushing, and cleaning up the cmux pane used to do it.

Use when David wants to create visual whiteboard graphics for YouTube videos using the tldraw SDK, OR inspect/extract content from an existing .tldr board file. Triggers on tldraw, whiteboard, visual diagram, canvas graphic, video visual, board design, "what's in this board", "read this column from the tldr".

Use when David asks about tasks, Todoist, to-do lists, priorities, inbox, or anything related to task management. Handles reading, creating, updating, moving, completing, and deleting Todoist tasks using David's specific setup and rules.

Use when David wants to manage his VPS servers and the AI agents running inside them — connecting, deploying, monitoring, restarting, and operating remote hosts and their agents. Triggers on VPS, server management, remote host, SSH into server, manage my servers, agents on the server.

How Pi accesses the web — search, fetch URLs/PDFs/YouTube/GitHub, and retrieve current information. Use whenever a task needs facts, docs, current/latest info, news, prices, versions, or content from a specific URL. Routes by the user's phrasing (web search vs extensive web research vs deep research). Backed by the globally-installed pi-web-access package.

Use when David is choosing the next YouTube video idea. Helps evaluate candidates, apply David's principles, and reach a decision.

Use whenever the user needs the transcript of a YouTube video — fetching, extracting, downloading, or pulling captions/subtitles/transcript text from a YouTube URL. Triggers on "get the transcript", "transcript of this video", "pull the captions", "download subtitles", "what does this YouTube video say".