Aider Getting Started: Terminal AI Coding With Per-Edit Git Commits
Install Aider, point it at a git repo, and run the see-diff/accept loop from your shell. No IDE, no browser, no lock-in. Updated June 2026.
Articles tagged with #Tutorial
Install Aider, point it at a git repo, and run the see-diff/accept loop from your shell. No IDE, no browser, no lock-in. Updated June 2026.
The full ChatGPT shortcut list (Mac + Windows), verified June 2026, plus which seven actually save time and which to skip.
How to wire MCP servers into Claude Code with claude mcp add: the three transports, the three scopes (local/project/user), and which servers to install first (June 2026).
The four Cursor rule types, .cursor/rules/*.mdc vs legacy .cursorrules vs AGENTS.md, and how to scope rules so the model obeys them. Current to June 2026.
Set up Windsurf — rebranded Devin Desktop in June 2026 — install, sign in, write a rules file, and ship your first reviewed multi-file Cascade diff.
Prep a 1-on-1 with AI in 10 minutes — agenda, topics worth raising, and a same-day follow-up. With the privacy settings to flip first.
Design AI album covers that survive Spotify's 64-pixel thumbnail — composition, palette, type, and the iteration loop that gets you to a final in under an hour.
A 90-minute Amazon listing workflow with AI — title, 5 bullets, description, and A+ modules, with the exact 2026 character and byte limits.
Teardown five competitors in 30 minutes with Perplexity and Deep Research, plus a structured comparison sheet — with verified June 2026 pricing and tool limits.
Plan a 12-episode content series with AI in one 2-hour session — arc, per-episode beats, hooks, repurposing, and a cadence that survives to episode 12.
Use AI plus Search Console data to find decaying posts, draft targeted refreshes, and recover rankings. Real numbers, tools, and a 30-minute workflow.
Script-first workflow for a 60-second AI explainer: 3-beat script, 6-8 shot storyboard, Sora 2 or Veo 3.1 visuals, ElevenLabs voice. With June 2026 tool pricing.
Build a fantasy character bible in Midjourney V7 and Flux 2: turnaround sheet, environment splash, and action variant on one face (June 2026).
Build a 6-outfit fashion lookbook on one consistent AI model with a frozen brand palette, using Midjourney Omni Reference or Nano Banana Pro (June 2026).
Use AI to find archives, decode old handwriting, and translate period languages — verified against the scan, never as a substitute for reading the document yourself.
Export your inlink graph, let AI score bridge candidates by topical fit and authority, and ship the 20 highest-value internal links in one PR.
Scope a per-locale launch with AI in 90 minutes: hreflang return tags, Intl.NumberFormat currency, and copy-paste verification commands.
Detect cannibalizing article pairs with Search Console data + AI, then decide merge vs differentiate per pair. Real thresholds, prompts, and timing.
A four-block AI workflow for LinkedIn posts that read like a sharp peer, not a motivational bot — with the algorithm math and a first-hour comment plan.
Build a defensible TAM/SAM/SOM with AI doing the legwork — plus a triangulation step that catches the made-up numbers. Tools, pricing, and a worked example.
Make a 30-second AI music video where every cut lands on the beat: a Suno v5 track, fixed-duration Sora 2 or Kling shots, trimmed to a bar grid.
Draft quarterly OKRs with AI in 90 minutes, then run a 15-minute Monday check-in that catches drift by week 4 instead of week 8.
Draft, pressure-test, and review personal OKRs with AI in 90 minutes plus a 15-minute weekly check-in. Tool picks and prompts current to June 2026.
Rewrite Slack drafts with AI so they read direct, kind, and move the thread forward. Tools, prompts, and what Slack AI can and can't do (June 2026).
Run a PRISMA-grade systematic review with AI as a screening and extraction layer — never as a citation generator. Tools, prompts, and disclosure rules for June 2026.
Write TikTok scripts with ChatGPT or Claude using the hook-reveal-CTA arc, timed to the 21-34s completion sweet spot and the data that drives FYP distribution in 2026.
Build a 45-second AI trailer with a real tension arc (setup, escalation, button) using Veo 3.1 and Kling 3.0, with June 2026 pricing and shot counts.
Pick the right tasks, run it in a sandbox, and verify every few steps. A reproducible Computer Use workflow with June 2026 models, pricing, and OSWorld numbers.
Dictate into Claude on your phone and arrive home with structured output, not a transcript graveyard. Step-by-step for June 2026.
Skills look like a toggle but behave like a matcher. Here is exactly how Claude reads a SKILL.md description and decides to fire, with the limits and fixes.
A shared Claude Project rots in three months without an owner. Here is the structure, file limits, and quarterly review that keep a team knowledge base answering questions past launch week.
Side-by-side on PRDs, JIRA grooming, and doc cleanup — with current pricing, models, and a 90-minute test you can run on your own week of work.
Gemini 3.1 Pro Deep Research can read 30+ sources in one run, and Deep Research Max can run overnight. Here is how to scope it so the depth pays off.
Gemini's Help me schedule and Add to calendar tools turn a Gmail thread into a booked meeting. Here is the exact setup, the real limits (20 guests, desktop-only, English-only), and the guardrails that keep it reliable.
Canvas puts a live doc beside Gemini chat. Use Select & ask plus version history and you stop pasting drafts back and forth.
Gemini in Sheets writes pivot tables from a plain-English prompt and an Apply button. Here's the verify-before-you-ship workflow, plus exact access steps and plans as of June 2026.
Run Suno v5.5 in structured batches: one brief, single-variable sweeps, a 5-point grid, two-pass culling. Get one finished track from 30 takes without decision fatigue.
Run Suno v5.5 and Midjourney V7 off one shared brief so your track and its 3000x3000 cover feel like a single release, not two unrelated assets.
Export Suno tracks as time-aligned 12-track WAV stems for real mixing and remix work — which plan you need, what bleeds through, and how to fix it as of June 2026.
A verified Codex CLI keyboard shortcut reference: cancel and interrupt keys, newline with Ctrl+J, Ctrl+G external editor, Ctrl+R history search, Esc Esc to edit, Shift+Tab approval modes, and terminal fixes.
How Codex Agent Skills work in June 2026: what a SKILL.md is, the exact .agents/skills/ load order, how /skills and $name invocation differ, and skill vs AGENTS.md vs MCP vs plugin.
Every Claude Code keyboard shortcut verified against the official docs (June 2026): the 6 permission modes, the Esc/Esc rewind, readline editing, custom keybindings, vim mode, and terminal fixes.
How Claude Code skills work as of June 2026: the SKILL.md frontmatter reference, auto-trigger vs /command invocation, dynamic context injection, subagent forks, and team sharing via plugins.
A current, hands-on reference to Claude Code's built-in slash commands as of June 2026: session lifecycle, /rewind checkpoints, /usage cost, /plan and /effort, subagent / hook / MCP extensions, plus what changed (/vim and /pr-comments are gone).
Outline tutorials that learners actually finish. 12 prompt templates for level-appropriate steps, prerequisites, success checks, and reuse.
Autocomplete predicts your next edit in ~100ms from local context; agents plan across files over 5-120s. A decision rule that maps task shape to the right mode, with current tool tiers.
A reusable workflow to generate testable ad variants for Meta, TikTok, and Google — built for Meta's 2026 Andromeda creative-diversity era.
Use Claude Code, Codex, or Cursor agents to pre-review your PRs before a human sees them. Exact prompt, loop, and tool pricing for June 2026.
Run an AI pre-review against the exact buckets Google AdSense rejects on — thin content, missing policy pages, broken navigation, edge-policy topics — before you burn a reapply window.
Prompt patterns for splash, settings, onboarding, and empty-state backgrounds that sit behind your UI instead of fighting it — with exact 2026 sizes and tools.
Use a reasoning-grade AI as a structured devil's advocate to catch 3-5 real architecture flaws per design doc before you commit weeks of code.
A reproducible AI audit for Astro 5/6 content sites: catch duplicate slugs, missing translations, dead internal links, and draft leakage with one script wired into prebuild.
Astro-specific AI audit prompts for routing, sitemap, hreflang, output mode and RSS — verified against Astro 6 (June 2026).
A reproducible AI audit for React Native: re-renders, navigation memory, New Architecture native-module compat, list virtualization — a ranked hotspot list with file paths.
Use AI image models to explore 6 visual directions in an afternoon, then hand a designer real references — with the right tool and prompt for each step.
Use AI to distill brand voice from 5-10 real samples into a 1-2 page style doc you can load into a Custom GPT or Claude Project as the system voice.
A 30-minute monthly workflow: lychee or linkinator finds the dead URLs, AI clusters them by root cause and proposes fixes in priority order. Exact commands and flags included.
A 25-minute Sunday AI workflow that scores meetings, blocks deep work, and drafts the async-shift messages — plus when to upgrade to Motion or Reclaim.
A 5-step AI workflow that turns a one-paragraph brief into 12 differentiated campaign concepts — each scored on a 3-axis rubric, ranked, and ready for stakeholder review.
Category and tag pages eat crawl budget but rarely rank. This AI workflow finds duplicate intros, weak internal links, missing descriptions, and JS-only pagination — then ships fixes.
A repeatable image-to-video workflow for AI character clips that move naturally. Constrain the motion, lock the camera, batch-generate. Verified June 2026.
Write camera language that makes Veo 3.1, Sora 2, and Kling 3.0 clips feel cinematic — named moves, start/end framing, one intent per shot.
A reproducible 4-pass workflow to verify every citation in an AI-assisted draft — catch fabricated sources, mismatched quotes, and chimera references before publish.
A five-question workflow for Claude Code, Cursor, or ChatGPT that maps an unfamiliar repo in an afternoon — with verification steps so you don't ship a tour built on hallucinations.
AI coding quality is mostly a context problem. Here is how to feed Claude, Cursor, and Codex the right context — and cut the rest — as of June 2026.
Keep the same AI character across 20 scenes using a canonical reference, a frozen trait list, and the 2026 reference features in Midjourney V8, Nano Banana Pro, and ChatGPT.
Use AI to pick pillars, generate clusters, verify demand, and build a 90-day calendar with an internal-link map.
A lean 7-piece AI content marketing stack with June 2026 pricing — ideation, drafting, editing, scheduling, analytics, repurposing, and assets, plus what to cancel.
Turn one long-form piece (blog, podcast, YouTube) into 10+ platform-tailored posts in 60-90 minutes, with verified 2026 platform limits and tool picks.
A 6-week AI-assisted workflow to build a distinct creator brand: audit your voice, sharpen your POV, lock signature topics, and ship a content engine.
A reproducible 7-step workflow for debugging with AI that finds root cause fast instead of spiraling into 20 wrong guesses. Includes a copy-ready prompt and tool picks for June 2026.
Upgrade a dependency that is several majors behind using Claude Code or Cursor, without subtle build breaks. CHANGELOG-mapped diffs, exact commands, June 2026.
A daily 15-minute AI email triage workflow with the right 2026 tools, a voice file, and replies that still sound like you.
A 20-minute AI workflow that compresses a 30-page report or 90-minute meeting into a 1-page exec summary built to survive the C-suite skim. Tool picks current as of June 2026.
A reusable 3-minute workflow to fact-check any claim with Perplexity plus a manual source check — built on the SIFT method, with current 2026 tool settings.
A 10-minute AI audit of firebase.json before you ship: rewrite order, function regions, cache headers, and preview-channel testing.
An AI-assisted playbook for content sites: information architecture, drafting, internal links, schema, and an AdSense-safe quality bar (June 2026).
Draft a 10-15 slide deck with AI in 30 minutes: outline, slide schema, speaker notes, plus which tools to use as of June 2026.
Use an LLM to audit hreflang on a multi-locale site: missing return tags, wrong codes, slug mismatches. Real prompt, real tools, June 2026.
One reference table for every place an AI image ships — blog OG 1.91:1, Instagram 4:5, Stories 9:16, Pinterest 2:3 — plus how GPT Image 2, Midjourney, and Gemini set ratios in 2026.
The six parts every usable AI image prompt needs, the three traps that flatten output, plus a tool comparison and copy-ready template for June 2026.
Stop rewriting the same image prompts every Monday. Build a 10-template library that turns recurring image work into fill-a-variable-and-ship — with model-specific syntax for Midjourney, GPT Image 2, and Imagen 4.
Your image series starts looking like six different artists made it. Here is the anchor-and-lock workflow that pulls a Midjourney, SDXL, or Flux set back to one voice.
Turn ChatGPT, Gemini, or Claude Deep Research into a defensible industry brief in 90 minutes. Engine limits, prompts, and the spot-check loop (June 2026).
An 8-step AI workflow to clean a sprawling Notion / Confluence / Google Sites wiki: export, batch-dedupe, triage, merge to canonical, archive, tag, reindex. Verified June 2026.
Draft a full landing page with AI — hero, benefits, social proof, FAQ, CTA — then ship it on Carrd or Framer and A/B test the result.
A 90-minute AI workflow that drafts every launch-day asset: announcement, landing hero, email, press blurb, demo script, and FAQ — with copy-paste prompts and 2026 length rules.
A repeatable 5-minute workflow to turn any Zoom, Meet, or Teams transcript into clean decisions and owned action items with AI — with real tool pricing (June 2026).
Draft meta descriptions in batches of 10 with AI, spot-check 30% for hallucinations, ship in one PR. Realistic CTR lift on thin pages: 10-20% as of June 2026.
A reproducible prompt workflow for framework and language migrations — reference file, pattern propagation, codemod, official-checklist review. Tools and model picks for June 2026.
AI agents in monorepos either ship clean PRs or melt down across 6 packages. Here is the AGENTS.md + scoping + two-phase commit workflow that keeps them contained, as of June 2026.
Synthesize across 5-50 sources with NotebookLM or Claude/ChatGPT Projects: a four-question workflow (consensus, conflict, gap, recency) with traceable citations, in one working day.
A 60-minute weekly newsletter workflow with AI — ideation, draft, subject-line A/B, send — plus current ESP pricing and a one-week feedback loop.
A 3-pass system for reading research papers with AI — triage, structured summary, and a manual deep read where AI only clarifies. With prompts, tools, and a citation-ready note format.
Build realistic AI portraits with a 5-slot prompt (subject, expression, wardrobe, lighting, lens) plus the exact tools, lenses, and texture words that beat the plastic-skin default. Updated June 2026.
The hybrid poster workflow for June 2026: pick the right AI image model, plan negative space, then set headline type in Figma or Affinity for crisp print output.
Wire an AI agent into your git workflow to scan the staged diff before you commit. Catches leftover debug code, leaked secrets, and null bugs in under a minute.
A four-angle workflow for product ad images that sell — with the right model (Nano Banana Pro, Seedream 5, Midjourney V7), real platform specs, and brand-safety guardrails.
A real workflow for product b-roll and short commercials with AI video tools (Runway, Kling, Veo) — and how to hide the AI tells, as of June 2026.
A photographer's 5-slot prompt scaffold, the camera vocabulary that signals real photography, and which June-2026 model (Nano Banana Pro, GPT Image 2, Seedream 5.0) to pick per product.
A step-by-step AI workflow that converts a feature list into 3-5 customer stories using the StoryBrand and ABT frameworks — with exact prompts and model picks (June 2026).
The three places AI portraits fail at full resolution — plus the exact prompt vocabulary, model picks, and retouch pass that fix each.
A safe AI refactoring workflow: green tests, scoped goal, plan mode, per-step git diff, and clean rollback — with the exact Claude Code and Cursor steps.
A repeatable 6-stage AI research workflow with exact 2026 tool limits — Perplexity, NotebookLM and Deep Research (ChatGPT/Gemini/Claude) chained so the output survives a citation check.
Cursor, Claude Code, and Copilot all ship checkpoints now — but they're local-only and miss bash edits. Here's the four-step git discipline that actually keeps every AI change reversible.
A weekly 30-minute AI workflow for Search Console: Page Indexing, Core Web Vitals (with 2026 thresholds), HTTPS, and Manual Actions — sort blocker vs. noise fast.
Vertical, fast, hook-first. Platform-specific AI video prompt templates for Douyin, TikTok, Reels, and Shorts, with the right tool and length for each.
A repeatable workflow to draft Reels / Shorts / TikTok scripts with ChatGPT, Claude, or Gemini that win the 3-second hold.
A repeatable AI workflow to draft a 4-week social calendar — pillars balanced, hooks pre-written, 30% left open for reactive posts. With June 2026 tool prices and posting-cadence data.
A 4-step AI workflow that interviews you, drafts the SOP, fact-checks it against live tool screens, and ships a usable v1 in under 90 minutes. With a 2026 tool comparison.
Turn a one-page spec into shipped code with Spec Kit, Kiro, or Claude Code plan mode — and skip the half-built-feature trap. Updated June 2026.
A 15-minute Friday AI workflow with copy-ready prompts that produces a status update your skip-level forwards — without sounding machine-written.
Turn a spreadsheet into a 4-sentence executive narrative with AI. Two paths (paste a pivot vs. upload to a code-running tool), exact 2026 tool limits, and a number-by-number fact-check.
A repeatable AI workflow to decompose any deadline goal into milestone → week → day, with ready-to-paste prompts and the right tool for each step (June 2026).
Skip the generic 80-point list. Have AI build a 30-60 item technical SEO checklist for your stack (Astro / Next / Hugo / WordPress), each item with a verification command, in 20 minutes.
AI-written tests often pass while testing nothing. This adversarial workflow plus mutation testing gives you real coverage, with the 2026 tooling that proves it.
A 3-pass AI workflow that scores every page for thin-content risk, surfaces the worst offenders, and assigns one action each: expand, merge, noindex, or delete.
Exact aspect ratios, resolutions, and safe-zone pixels for TikTok, Reels, YouTube, and ads — plus which AI video tools generate each ratio natively as of June 2026.
Turn a finished script into 8-12 well-paced AI shots: break it into shot-level prompts, decide literal vs evocative per shot, then cut audio first. Tools, clip limits and costs as of June 2026.
Subject morphs, background warps, camera glitches? Diagnose 1 of 5 drift types, change 1 variable, and get a clean clip in 4-6 tries instead of 20.
The 6 parts every AI video prompt needs in 2026 — written for Veo 3.1, Kling 3.0, Runway Gen-4.5, and Sora 2. Real structure, real numbers.
A repeatable AI video workflow for Meta, TikTok, and YouTube Shorts ads: which tools, exact costs per clip, and how to test 5-8 hooks without burning budget.
Storyboard before you generate. With Veo, Runway, and Kling capped at 5-8s clips, planning multi-shot sequences first beats prompt-and-pray.
Why a 5-clip AI sequence looks like 5 productions, and the exact reference-image + style-bible workflow that makes Runway, Kling, and Veo clips feel like one piece.
A daily 90-second capture habit plus a Friday AI prompt turns raw notes into a tight 3-2-1-3 weekly report. Tool picks, model versions, and the exact prompts, current as of June 2026.
A 30-minute AI audit before launch: run three focused prompts (security, performance, UX) against the OWASP Top 10:2025 and get a triaged, diff-ready fix list.
A 20-minute AI bug-audit habit that beats 'review this code': category-grouped findings, a failing test per bug, and likelihood × blast-radius triage. Models and tools current to June 2026.
Set up ChatGPT in 30 minutes: pick the right plan, make the one settings change that matters, and run three prompt patterns that actually work.
A sparring-partner workflow for ChatGPT — diverge to 10 angles, attack the best 3, synthesize. With the right model picker and prompts for June 2026.
ChatGPT is not Claude Code, but used right it still ships real features. A GPT-5.5 workflow with Projects, code blocks, and review steps that hold up on actual files.
GPT-5.5 Instant for chat, Thinking for hard math and code, Pro for the rare unverifiable task. A decision sheet with real limits and context windows.
A Custom GPT is a saved system message plus up to 20 knowledge files. Build one for a workflow you run weekly.
Upload a CSV to ChatGPT, run real Python, and get a chart you can trust — the profile-first workflow, current limits, and three sanity checks.
ChatGPT runs real Python on your CSV. Here is a step-by-step workflow to get correct numbers, not just good-looking charts, with June 2026 limits and settings.
A repeatable workflow that makes ChatGPT cite page numbers and pull real figures instead of vague summaries — with current file-size and upload limits.
Use ChatGPT for regex, scripts, SQL, and debugging in the browser. Exact GPT-5.5 limits, when to use code blocks vs Code Interpreter, and when to switch to Cursor.
How to get usable images from ChatGPT Images 2.0 (gpt-image-2): structured prompts, one-variable iteration, masked edits, and reference-image consistency.
A repeatable ChatGPT Images 2.0 workflow — Instant vs Thinking modes, real aspect ratios, plan limits, and the iteration loop that gets usable images, not lucky first tries.
Run a multi-week deliverable across many ChatGPT chats without losing the thread — using one Project, two files, and project-only memory.
ChatGPT now remembers you in two ways and rewrites those memories on its own. Here's how the June 2026 system works and how to keep it accurate.
Stop defaulting to one model. A per-task framework for ChatGPT's GPT-5.5 Instant, Thinking, and Pro — with real 2026 rate limits, context windows, and prices.
ChatGPT pins are capped at 3. Here is the archive-plus-Projects system that keeps a 500-chat sidebar findable.
A staged outline-first method that beats one-shot summaries, plus the exact 2026 file limits and when Claude or NotebookLM is the better tool.
Keep notes in Obsidian or Notion, use ChatGPT only as the retrieval and synthesis layer. Snapshot cadence, Project setup, privacy settings, and NotebookLM vs ChatGPT — all current for June 2026.
A disciplined Projects setup that holds context across weeks: file limits, project memory, instructions, and the decisions.md habit that actually works.
Four levers that fix most weak ChatGPT output, plus the GPT-5.5 habits to drop — outcome over process, decision rules over always/never.
Stop losing ChatGPT notes in chat history. A Projects-based research-notes loop that survives a multi-week literature review (June 2026).
Run a credible research project in ChatGPT — Deep Research mode, Projects for context, an editable research plan, and the citation spot-check loop that keeps output honest.
A real ChatGPT research workflow for briefs that survive scrutiny: outline, source by section, paste what you read, synthesize, and verify — plus when to use Deep Research vs Perplexity.
A two-pass ChatGPT workflow that tailors your resume to the job description and ATS keywords, then puts your real numbers back in — without the generic robot voice recruiters skim past.
Three ways to use ChatGPT on Excel and Sheets — formula help, file-upload data analysis, and the in-sheet sidebar — with the verification steps that keep your data safe.
Turn ChatGPT into a tutor: Study Mode, active-recall quizzes, and Anki flashcards. Specific prompts plus 2026 plan limits for students.
Voice is faster than typing for 5 task shapes and worse for everything else. Tier limits, Advanced vs Standard, video sharing, and the exact line — as of June 2026.
ChatGPT Voice is fast but flaky. Here is exactly when it earns its keep, the June 2026 limits, and a walking-prep template you can copy.
How to use ChatGPT search (June 2026) to find current facts and verifiable URLs — exact toggles, prompt structure, and the source check most people skip.
ChatGPT search looks like a research shortcut. Here is the June 2026 workflow that makes it actually one — with a 90-second verification routine.
A four-stage writing workflow with ChatGPT (GPT-5.5, June 2026): pre-writing, your draft, AI as structural editor, cautious polish — plus copy-ready prompts that beat "rewrite this."
Pick by task, not feature list: NotebookLM for sources you upload, Perplexity for fast cited web answers, Deep Research (ChatGPT/Gemini/Claude) for long agentic briefs. With June 2026 prices and limits.
A categorize-first workflow for using Claude (Sonnet 4.6 / Opus 4.7) to turn 10+ messy inputs into a defensible recommendation, with real prompts and 2026 limits.
Treat an Artifact as a real mini-app: spec it, iterate by diff, add the new persistent storage and Claude API binding, then publish a share link. Field-tested, June 2026.
How Claude Artifacts work in 2026 — types, the AI-powered app billing trick, sandbox limits, and a tested build-and-ship workflow.
Set up Claude the right way: pick a plan, build your first Project, and run a long-document task where Claude actually beats ChatGPT. Verified June 2026.
Claude in Chrome can read and act on your tabs. Here is the default-off workflow, the exact permission settings, and the security facts you need as of June 2026.
Install Claude Code, write a CLAUDE.md, run Plan Mode, and ship your first reviewed refactor. Exact commands, plan/cost details, and the beginner mistakes to skip.
Install Claude Code, write a CLAUDE.md, and run safe agentic edits. Pricing, models, permission modes, and a beginner workflow — current as of June 2026.
CLAUDE.md is the instruction file Claude Code loads every session. What to put in it, where to place it, and how it loads — verified against the June 2026 docs.
Set up Claude Code on a real codebase the right way — install, CLAUDE.md, permissions, and a safe first task — in about 30 minutes (June 2026).
A reusable Claude Code plan-mode prompt and a two-hour workflow that ships three reviewed SEO fixes on your content site.
A daily Claude Code workflow that survives past week one: the official Explore → Plan → Implement → Commit loop, plan mode, /clear and /compact context hygiene, subagents, and CLAUDE.md upkeep — with June 2026 plan and model specifics.
Set up Claude's Gmail, Drive, and custom MCP connectors the safe way: narrow OAuth scope, monthly audits, and exactly where to revoke. Tested June 2026.
Claude Connectors let the Claude app search, read, and now write to Drive, Notion, and Slack mid-conversation. What each does, plan limits, scoping, and where they break.
Claude's file handling beats most LLMs — if you upload right. Exact limits, a grounded prompt sequence, and citation checks for contracts, papers, and reports.
Claude file uploads explained: 30 MB and 20-file chat caps, 100-page PDF analysis, Project knowledge with RAG, and prompts that actually retrieve. June 2026.
Claude connects to GitHub three different ways. Here's which one to use for code research, terminal commits, and automated PR review — with the exact setup as of June 2026.
A page-cited workflow for processing 100-500 page PDFs in Claude (June 2026 limits) without splitting, mis-paraphrasing, or hallucinated sections.
Claude.ai caps PDFs at 100 pages, so a 200-page report needs a split-and-merge workflow. Here is the exact step-by-step for deep, cited analysis (June 2026).
Set up Claude Projects so the 500K context and RAG actually help — curated files, a decisions log, and a two-week prune cycle.
How Claude Projects share state, what goes in Knowledge vs Instructions, and a setup that survives months of work.
Claude Opus 4.7 reads instructions more literally than any model before it. Here is how to write prompts it gets right on the first try.
Upload, outline, multi-angle explanation, quiz, next-day recall — a deliberate-practice loop with Claude Sonnet 4.6 and Learning mode that turns passive reading into real retention.
Real upload limits, context windows, and a task-by-task verdict for 100+ page docs in Claude vs ChatGPT.
A repeatable Claude writing workflow: the five-part prompt, the Styles feature, and how to kill the AI tell. Updated June 2026.
How OpenAI Codex runs sandboxed cloud and CLI coding tasks in 2026: which ChatGPT plan you need, the GPT-5.5 default, the task-spec format that produces reviewable PRs, and the mistakes that bite first-time users.
Set up OpenAI Codex as a pre-reviewer with the /review command, @codex review on GitHub, and AGENTS.md guidelines that produce P0/P1 findings.
Use Codex CLI to spot-check sitemap correctness, freshness, and hreflang coverage against your real build output.
Goal + constraints + acceptance + hand-off. Skip a block and the AI guesses. The four-part template, with 2026 plan-mode and AGENTS.md specifics.
Cursor's real learning curve isn't install — it's knowing when to use Tab, Inline edit (Cmd-K), Composer (Cmd-I), Chat (Cmd-L), or Agent mode. Five surfaces, four real edits, with June 2026 pricing and models.
How Cursor's codebase index actually works, the .cursorignore vs .cursorindexingignore split, and the tuning loop that stops wrong-file references — verified June 2026.
A 30-page audit dumped into chat produces 'did some refactoring.' Use this chunk-summarize-prioritize pattern to turn long reports into a brief Claude Code, Codex, or Cursor can actually execute.
The five Gemini-on-Android tasks that beat phone typing — with the exact gestures, free Gemini Live setup, and what to leave for your laptop. Updated June 2026.
Start with Gemini free, run a real Gmail + Docs task in 30 minutes, and learn exactly when Google AI Pro ($19.99/mo) is worth it.
Gemini's brainstorming edge isn't more ideas — it's pulling your Drive Docs, decks, and Sheets in via @-mentions to frame against real context. A repeatable 30-minute session for Gemini 3.1 Pro.
Gemini 3.1 Pro is a real coding tool now. Honest guide to its 4 genuine wins, the Antigravity CLI migration on June 18, and when to switch to Claude or ChatGPT.
A verify-first workflow for Gemini Deep Research on Gemini 3.1 Pro: edit the plan, audit every source, and keep only what survives. Limits and pricing as of June 2026.
Gemini in Docs is best used in five surgical spots — outline, expand, tighten, fact-check, format. Exact prompts, the new bottom-bar UI, and what each plan unlocks as of June 2026.
A 60-page PDF, 20 minutes, and a structured, page-referenced summary you can spot-check. The exact Gemini prompt sequence, file limits, and verification habits that keep you out of trouble.
Which Gemini-in-Gmail features are free, which need Google AI Pro, and the exact prompts to clear 30-100 emails a day without sounding like a robot.
Chain Gemini's in-line Gmail tools and the Docs side panel into one read-draft-polish-reply loop, with current 2026 features, plans, and limits.
Gemini's real edge is native access to Gmail, Docs, Drive, Sheets, and Calendar. Here is the surface-by-surface playbook, with Gems, as of June 2026.
Gemini's image generator now renders legible text and 4K detail. The exact model picker, free vs paid limits, and a five-slot prompt recipe for usable posters on the first try.
The Gemini limits people hit first — the 32k free context window, region-gated Deep Research, blocked real-people images, and weaker Chinese — with the exact tier and workaround for each (June 2026).
A repeatable Gemini workflow that turns Calendar, Drive, and Gmail into a one-page meeting brief in about 8 minutes, using Connected Apps and a saved Gem.
A reproducible Gemini PDF workflow with exact file limits, the structure-first prompt sequence, and how to verify every number Gemini gives you.
Gemini Deep Research (Gemini 3.1 Pro) wins on fresh web data + Workspace export. Here's the exact 2026 workflow that produces a defensible brief in about 45 minutes.
How to use the Ask Gemini side panel and Fill with Gemini to write formulas, build pivots, and clean data — plus the verification habits that stop silent errors.
A retention-first Gemini study workflow using Guided Learning, native flashcards and quizzes, and NotebookLM — with exact steps as of June 2026.
How to set commit conventions, draw commit boundaries, and what NOT to let Claude Code, Cursor, or Codex commit. With real config.
A 200-line AI diff that compiles isn't safe. Here's the exact reading order, git commands, and tooling a senior engineer uses to catch silent renames, lost branches, and fake test fixes.
Turn one still into clean motion without the product warping or a third hand appearing. Tool-by-tool settings for Runway Gen-4.5, Kling 2.6, and Luma Ray3 (June 2026).
A hands-on Kling 3.0 image-to-video workflow: mode and credit picks, Motion Brush, prompt patterns, and where it still breaks. Updated June 2026.
Get usable Midjourney images in 30 minutes: the one-sentence prompt formula, the 5 parameters that matter (--ar, --raw, --s, --sref, --oref), Draft Mode, and the iteration loop pros use — updated for V7 and V8.1.
Run Cursor, Claude Code, and Codex in one repo using git worktrees, AGENTS.md, and PR-based merging — the 2026 playbook that prevents collisions.
Edit photos by prompt with Nano Banana (Gemini 2.5/3 Flash & Pro Image). Pick the right model, write keep-the-rest edits, and avoid the 6 common mistakes.
A practical NotebookLM walkthrough for June 2026: real source limits, Gemini 3.1 Pro, Audio and Video Overviews, plus the questions that actually surface insight.
Perplexity sits between search and chat: cited answers with web freshness. Free vs Pro limits, models, and a real research workflow (June 2026).
Polite CLAUDE.md rules don't stop an agent from touching .env or prod config. The enforced layers that do: deny rules, ignore files, sandboxing, hooks, branch protection.
A task-by-task decision sheet for picking the right AI assistant for office work — meetings, email, docs, slides, sheets — with June 2026 pricing and models.
A practical Runway tutorial for June 2026: Gen-4.5 vs Gen-4 Turbo, real credit costs per second, project setup, prompt patterns, and an export checklist.
A repeatable AI prompt chain that audits titles, descriptions, schema, and internal links page by page - with the verification step that stops AI from lying about tags it can't see.
Make your first AI video in 15 minutes. How Sora and Veo access actually works in June 2026, the prompt structure that works, clip-length limits, and which model fits realism vs. surreal.
Match palette, lens, and lighting across 10+ AI images with a style anchor, Midjourney --sref, Omni Reference, and Flux LoRAs (June 2026).
Make your first Suno song in 15 minutes - Simple vs Custom Mode, style and lyrics prompts, structure tags, and 2026 pricing/credits.
A tight Suno workflow for 5-15s brand jingles: exact prompt structure, length and BPM specs, Personas for a consistent voice, and the June 2026 commercial-rights checks to run before you ship.
Suno v5.5 does real guofeng well if you pull three levers: named Chinese instruments, explicit pentatonic mode, and Mandarin lyrics in classical-couplet rhythm. Full prompt template + credit budget inside.
The chorus is the part listeners remember. Here is how to push Suno v5.5 past generic hooks with structure tags and a written hook.
Suno commercial rights hinge on your plan when the song was generated. Plus the post-Warner ownership change and what each platform now demands.
Suno Cover re-performs your audio in a new genre on the v5.5 model — but it never grants rights you don't already hold. Free uploads cap at 60s, Pro/Premier at 8 min. Here's the clean workflow plus the licensing line you can't cross.
Set the cut point, write a continuation prompt that adds new material, and stitch parts with Get Whole Song — a tested Suno v5 Extend workflow (June 2026).
A repeatable Suno v5.5 workflow for a 7-15 second brand jingle: write a 3-5 syllable hook, prompt for length and brand sound, generate 6-10 takes, and ship an intro/sting/outro set.
Suno clips drift if you just keep hitting Extend. Here is the structured way to reach a full 3-4 minute song on Suno v5.5 and keep its identity.
Write to a real pop structure with the bracket tags Suno v5.5 actually reads. Line counts, section jobs, and a copy-ready template.
A Suno style prompt is a recipe card, not an English sentence. Here is the field-by-field structure that actually changes the output on Suno v5.5.
Suno raps well if you stop treating it like a singing model. A genre-specific workflow with copy-ready trap and boom-bap prompt templates, tested on Suno v5.5.
Generate royalty-free short-video BGM with Suno v5.5: mood-first prompts, a 15-30s loop trim, and the per-platform loudness targets that actually matter in 2026.
The ordered Style-of-Music block that gets consistent Suno v5.5 songs: one genre, explicit BPM, 2-4 instruments, mood, production cues.
Make Suno sing the right language and accent. Model picker, native-script lyrics, accent tags, and a tonal-language fix — current as of June 2026.