$ claude -n maximum-output · june 2026 · v1.1

Maximum Output
with Claude Code

A developer's field guide to running AI like a system, not a chat box - persistent context, planning gates, controlled execution, and verification.

scroll to read  ·  P prints the full guide  ·  click the pipeline rail to jump

01 / PREMISE · why this exists

Same model. 10× the output.

Everyone here already uses AI. Output still varies wildly between developers - same tool, same plan, same model. The difference is the operating system you build around the model: context, session discipline, a research-first planning gate, and human verification.

One principle drives the whole workflow: errors get more expensive the later you catch them. Every step in this deck exists to push decisions earlier - where they're cheap - and every step names the exact cost of skipping it.

minutes
WRONG PLAN
hours
WRONG CODE
days
WRONG MERGE

Cost of catching the same mistake at each phase.

02 / OVERVIEW · the system at a glance

Eight stages. Two human gates.

ONE-TIMEContext
CLAUDE.md + ai-context
EVERY TASKSession
named · one task · right effort
EVERY TASKResearch
plan mode, docs-first
HUMAN GATEPlan review
visual HTML plan
EVERY TASKBuild
subagents, phase diffs
HUMAN GATETest
manual run + document bugs
EVERY TASKCommit
logical groups
EVERY TASKReview
machine, then human

highlighted = human gate. Everything else can be delegated to the agent. These two cannot - and most AI-assisted failures trace back to treating one of them as optional.

The next nine slides walk the pipeline left to right. Each one ends with the same box: what the step saves you from.

STAGE 01 / CONTEXT · one-time setup, permanent payoff

The repo's brain: CLAUDE.md

An agent without context is a very fast intern with amnesia. Claude Code auto-loads CLAUDE.md into every session - so the project's brain goes into a file once, not into prompts forever. Committed to git: every developer's agent gets it.

  • 10-second summary - what this repo is, and is not
  • Architecture map - where things live
  • Exact commands - build · test · lint · run. Never let it guess pytest vs npm test
  • Conventions - naming, errors, logging, "we use X, not Y"
  • Git rules - branches, commit format, PR checklist
  • Known gotchas + pointers to docs/ai-context/
# bootstrap, then prune hard - keep ~100–150 lines
$ /init
# link out to detail docs, don't paste them
# provider-agnostic - same brain for Cursor/Codex/Gemini:
$ ln -s CLAUDE.md AGENTS.md

Every line is context spent before work begins. Lean file, linked depth.

WHAT IT SAVES YOU FROM

Re-explaining the repo every session (10 min × every dev × every day). Wrong build commands. Five teammates getting five different conventions from the same model. New joiners' agents starting from zero.

STAGE 01 / CONTEXT · the other two layers

Personal layer + depth on demand

~/.claude/CLAUDE.md

Your private layer

User-level memory, loads in all your projects on top of the repo file: response style ("direct, no fluff, ask before assuming"), personal workflow rules, machine paths. Project-personal notes that don't belong in git go to local memory files - not the shared file.

docs/ai-context/

Depth on demand

Deep docs the agent reads when a task needs them: architecture decision records, domain vocabulary, API contracts, integration quirks. CLAUDE.md stays lean; depth stays one read away. Exactly the pattern we already run in production.

WHAT IT SAVES YOU FROM

Restating your preferences every session. The team file turning into one developer's diary. And the 600-line CLAUDE.md that eats half the context window before the task even starts.

STAGE 02 / SESSION · every task

One session = one task, named at birth

$ claude -n payment-webhook-retry # name at start
> /rename # or (auto-)name it later
> /model · /effort # firepower to match the task
> /context # where is the window going?
> /clear # new task = clean slate
$ claude -r "payment-webhook-retry" # a month later

One task per session

The window is the agent's working memory. Mix three tasks and all three reason over each other's leftovers - quality drops, tokens burn re-reading files.

Name it, resume it

A named session keeps everything - files read, decisions, your corrections. Resuming costs near zero; restarting re-pays the whole exploration bill.

Right effort per task

Planning and review deserve max reasoning; renaming a variable doesn't. Strongest model + high effort for design - faster model for mechanical edits.

WHAT IT SAVES YOU FROM

Paying twice to re-read the same files. Yesterday's refactor leaking into today's hotfix. "What was I doing?" archaeology. Token bills you can't explain.

STAGE 03 / RESEARCH · the highest-leverage step

Research-first planning

Plan mode (Shift+Tab · /plan) makes the session read-only - the agent can explore the codebase and the web but can't touch files. Then the key move, asked explicitly:

"Research first. Check current official docs and recent best practice
 for <problem> online before proposing anything. Cite sources.
 Then write the plan."
  • Training data ages from day one - your dependencies don't stop moving
  • Surfaces today's recommended API, next month's deprecation, the library that already solved it - with citations, in minutes
  • A human doing the same: ~2 hours of skimming, and still misses things
/plugin

Make it repeatable

Workflow plugins - superpowers, compound-engineering - turn brainstorm → research → plan → execute into a process instead of a mood. Install once; the discipline stops depending on how well you prompt today.

WHAT IT SAVES YOU FROM

Building on deprecated APIs. Reinventing a problem solved last quarter. Plans based on 2024 memory in a 2026 codebase. And the most expensive failure of all: confidently implementing the wrong approach.

STAGE 04 / HUMAN GATE Nº1 · the most underrated step

The visual plan gate

A 400-line plan as terminal text invites skimming - and skimmed approval is no approval. Before any code:

"Generate the approved plan as a local HTML page on a live server:
 phases, architecture & sequence diagrams, file-touch map, the
 reasoning behind each decision, risks and rollback."

Review it phase by phase like a design doc. Push back. Massage steps to fit the real requirement. Regenerate. Approve only when you'd defend every phase in code review. This is the cheapest moment in the entire lifecycle to change direction - and the page doubles as the design doc you attach to the PR.

WHAT IT SAVES YOU FROM

Rubber-stamping a wall of text. Design flaws discovered mid-implementation at 10× the fix cost. "That's not what I meant" after 2,000 lines. Writing the design doc twice.

STAGE 05 / BUILD · controlled execution

Build with subagents

Delegate implementation to subagents - isolated workers with their own context window, prompt and tool permissions (built-ins like Explore, custom ones in .claude/agents/). The main session stays the orchestrator: it holds the plan and your decisions; subagents burn their own context reading files and return distilled results.

PER PHASE

Work lands in slices

One plan phase at a time - never the whole feature in one shot.

git diff · /diff

You review each slice

While it's still small enough to actually read. Confirm, then release the next phase.

/rewind

Checkpoints

Approach went sideways? Roll code and conversation back together.

WHAT IT SAVES YOU FROM

Main-thread context bloat - reasoning quality decays as the window fills. One giant 2,000-line unreviewable diff at the end. The "agent ran 40 minutes unsupervised" surprise.

STAGE 06 / HUMAN GATE Nº2 · the non-negotiable

Manual testing + bug documentation

Tests passing, types checking, lint clean - that means the code agrees with itself, nothing more. The agent verifies what it can see; you verify what users feel. Run the feature in a real environment. Push the edge cases that come from knowing the business.

Found a bug? Fix it in-session while context is warm. Then the step almost everyone skips: document the bug class and feed it back into CLAUDE.md or docs/ai-context/ - "always handle timezone-naive datetimes at API boundaries." Every bug becomes a permanent rule the agent follows in all future sessions. That is the difference between a workflow that compounds and one that plateaus.

WHAT IT SAVES YOU FROM

"Tests pass, feature broken." Agent-written tests that only prove the agent's code agrees with itself. Paying for the same class of bug twice - or five times, across teammates.

STAGE 06 / EXAMPLE · from the field

One bug, one permanent rule

Manual testing on the timezone example from the previous slide turns up a real bug: a datetime crossing midnight UTC compares wrong against a naive local timestamp. Fixed in-session - then the rule goes into CLAUDE.md so no future session repeats it.

## Non-Obvious Rules **Service-layer contract.** Services call db.add() - never db.commit(). **Migrations are immutable.** Add a new migration for any schema change. **Tests need Docker running.** make up before make test.
## Non-Obvious Rules **Service-layer contract.** Services call db.add() - never db.commit(). **Migrations are immutable.** Add a new migration for any schema change. **Tests need Docker running.** make up before make test. Always handle timezone-naive datetimes at API boundaries - normalize to UTC on parse, never compare naive vs aware directly.
WHAT IT SAVES YOU FROM

The same bug class resurfacing in the next PR, written by the same agent, because nothing told it the rule changed. One sentence, committed once, applies to every session after - yours and every teammate's.

STAGE 07 / COMMIT · ship clean

Commits with intent

Group the work into logical commits - one concern each: schema change, business logic, tests, docs. The agent drafts messages from the actual diff (it read every line; its summaries are honest). You do the final read before push.

History is a debugging tool. A clean one is the difference between git revert and a 2 a.m. manual unpick.

WHAT IT SAVES YOU FROM

"final-fix-2-really" archaeology. Un-bisectable, un-revertable history. Reviewers forced to untangle five concerns inside one commit.

STAGE 08 / REVIEW · twice, in the right order

Machine review, then human review

PASS 1 · MACHINE

/code-review · /security-review

On the branch - or a Claude review wired into CI so every PR gets it automatically. Catches correctness bugs, security slips and convention drift before a human spends a minute.

PASS 2 · HUMAN

Judge the design, not the nits

With nits pre-cleared, humans do what only humans can: is this the right design? Does it actually solve the ticket? Will ops hate us in six months?

WHAT IT SAVES YOU FROM

Senior engineers spending review time on formatting. Security issues a bot catches in seconds. Review fatigue - the silent killer that turns "LGTM" into a rubber stamp.

10 / FORCE MULTIPLIERS · one-time investments, permanent returns

Beyond the core loop

.claude/commands/

Custom slash commands

A prompt you've written twice becomes a command: /fix-issue 123, /plan-html. Whole workflows in one keystroke, identical for the whole team.

.claude/settings.json

Hooks + deny rules

Deterministic guardrails: block destructive commands, auto-format after every edit. Unlike instructions, these always execute - the model can't "forget" them.

MCP SERVERS

Real data, not paraphrase

Wire Jira, Sentry, your DB, Figma. The agent reads the real ticket, the real stack trace, the real schema.

SKILLS

Packaged know-how

Deploy runbook, review checklist, domain rules - written once, every developer's agent gains the capability.

GIT WORKTREES

True parallelism

Multiple named sessions on separate branches simultaneously. Zero collision, full speed.

/rewind

Fearless experiments

Roll code and conversation back together when an approach dies. Checkpoints make boldness cheap.

11 / ANTI-PATTERNS · read this twice - all six happen in this building

What kills the gains

12 / CHEAT SHEET · one screen to remember

The whole system

STAGEMOVEWHERE / COMMANDWHAT IT SAVES
CONTEXTRepo brain, committedCLAUDE.md (+AGENTS.md) · /init → pruneRe-explaining repo · wrong commands
CONTEXTPrivate layer · deep docs~/.claude/CLAUDE.md · docs/ai-context/Restating style · bloated CLAUDE.md
SESSIONName · one task · resumeclaude -n · /clear · claude -r "name"Contamination · re-paying exploration
SESSIONRight firepower/model · /effort · /contextOverpaying easy, underthinking hard
RESEARCHRead-only + research/plan + "research first, cite sources"Deprecated APIs · wrong approach
PLAN GATEVisual plan review"plan as local HTML" → massage → approveDesign flaws at 10× fix cost
BUILDSubagents · phase diffs.claude/agents/ · git diff · /rewindContext bloat · mega-diffs
TESTManual run · document bugsfeed rules back into CLAUDE.mdRepeating the same bug class
COMMITLogical groupsagent drafts · final human checkUnrevertable history
REVIEWMachine pass, then human/code-review · /security-review · CINits eating senior review time
GUARDRAILSDeterministic rules.claude/settings.json - hooks, denyInstructions the model "forgets"
13 / CASE STUDY · the whole loop, once

One real ticket, start to finish

Same bug as the example two slides back, run through the full pipeline. Nothing here is exotic - it's the eight stages, in order, on a ticket small enough to fit a coffee break.

≈ 30 minutes · two human checkpoints · one bug class that can never recur.

13 / START TOMORROW · the 30-minute setup

Do this once. Then run one real ticket.

Honest expectation: the first full run feels slow - you're paying setup cost. The second feels normal. By the fifth, going back to "just write the code" feels reckless, because now you can see what each skipped step costs.

$ pipeline complete · exit 0

Docs: docs.claude.com/en/docs/claude-code - commands verified June 2026, Claude Code 2.x · Workflow & field notes: Anupam Kumar