9inches

Data Science | AI & Automation | $BTC since 2019 | Investor | 12 $ETH HODL until 2050

[9inches @0xnineinch]

This is f*cking insane. SpaceXAI engineer Nick Miller (ex-Cursor):

“stop making one bot do five different jobs

give each bot a role, connect them, and turn your chat window into a team”

In this 26-minute talk, he explains how Grok Bot can take over recurring work and coordinate specialist bots

→ 08:29 — the foundation: context, connectors, and a clear outcome for every bot

→ 13:03 — his Chief of Staff sends a 5 AM briefing with calendar events, inbox updates, and Slack priorities

→ 17:12 — split engineering work across frontend, QA, documentation, and research bots

→ 19:20 — bots can talk to each other: one coordinator brings in specialists to handle different parts of a task

the useful part isn’t collecting more bots

it’s giving each one a clear job and connecting their work

Save it and watch it today ↓

[9inches @0xnineinch]

Haggle Bot + Jev is a ruthless 24/7 AI team that could cut $100K+ from your software subscriptions and vendor bills

Elon Musk puts Haggle Bot’s potential at up to $6 trillion in annual savings if a 5% improvement in enterprise cost efficiency scales across the global economy

Jev handles the quick decisions: which bill needs investigation, which needs more data, and which can reuse a verified finding, so Haggle spends its time on expenses worth reviewing

connect your invoices, vendor contracts, and software usage records, then build a recurring workflow that catches wasted spend before the next renewal

if you’re running an agency, a software company, or a business juggling dozens of subscriptions, there’s a clear place to start

the full setup:

→ give Jev 5 possible next actions: reuse a verified result, request missing data, investigate a bill, prepare a vendor counter, or flag something for your review

→ let code compare each invoice with the previous version, checking prices, seat counts, renewal dates, and whether the same expense has already been investigated

→ send Haggle the contract, payment history, and actual usage, so it can look for unused seats, overlapping tools, unexpected price increases, and cheaper alternatives

→ have it turn each finding into a concrete proposal: which seats to remove, which subscription duplicates another tool, or what terms to ask the vendor to change

→ put cancellations and vendor messages in a review queue, with the supporting evidence, expected saving, and contract conditions attached

→ save every accepted finding with its inputs and outcome, so matching bills reuse the result while changed prices or usage trigger a fresh review

→ every morning, get one page showing new opportunities, decisions waiting on you, confirmed reductions, and what the AI workflow itself cost to run

run it for 7 days and compare the savings identified, the changes approved, and the money actually removed from future bills

keep the useful routines running around the clock, with a spending cap and clear rules for when the system needs you

save this and start with the subscriptions your team keeps renewing without checking

[9inches @0xnineinch]

this is the most UNDERRATED Grok Bot setup right now

Haggle Bot could make businesses 5%+ more cost-efficient, unlocking up to $6 trillion in annual savings if that improvement scaled across the global economy

take 40 paid seats, 10 active users, and $30 per seat — that’s $900/month or $10,800/year going toward licenses nobody uses, before checking a single other subscription

→ connect Ramp or upload your bills to build a vendor inventory
→ prioritize contracts renewing within the next 120 days
→ find unused seats, duplicate subscriptions, and compare 3+ alternatives per vendor
→ get a negotiation draft showing what you pay, the alternatives, and the terms to ask for

every savings estimate comes with supporting numbers
every outgoing message needs your approval

one forgotten subscription is easy to miss
20 subscriptions renewing unchecked can become a very expensive habit

find Haggle Bot on the Grok Bot Marketplace and start with your next renewal ↓

[9inches @0xnineinch]

Haggle Bot + Jev is a ruthless 24/7 AI team that could cut $100K+ from your software subscriptions and vendor bills

Elon Musk puts Haggle Bot’s potential at up to $6 trillion in annual savings if a 5% improvement in enterprise cost efficiency scales across the global economy

Jev handles the quick decisions: which bill needs investigation, which needs more data, and which can reuse a verified finding, so Haggle spends its time on expenses worth reviewing

connect your invoices, vendor contracts, and software usage records, then build a recurring workflow that catches wasted spend before the next renewal

if you’re running an agency, a software company, or a business juggling dozens of subscriptions, there’s a clear place to start

the full setup:

→ give Jev 5 possible next actions: reuse a verified result, request missing data, investigate a bill, prepare a vendor counter, or flag something for your review

→ let code compare each invoice with the previous version, checking prices, seat counts, renewal dates, and whether the same expense has already been investigated

→ send Haggle the contract, payment history, and actual usage, so it can look for unused seats, overlapping tools, unexpected price increases, and cheaper alternatives

→ have it turn each finding into a concrete proposal: which seats to remove, which subscription duplicates another tool, or what terms to ask the vendor to change

→ put cancellations and vendor messages in a review queue, with the supporting evidence, expected saving, and contract conditions attached

→ save every accepted finding with its inputs and outcome, so matching bills reuse the result while changed prices or usage trigger a fresh review

→ every morning, get one page showing new opportunities, decisions waiting on you, confirmed reductions, and what the AI workflow itself cost to run

run it for 7 days and compare the savings identified, the changes approved, and the money actually removed from future bills

keep the useful routines running around the clock, with a spending cap and clear rules for when the system needs you

save this and start with the subscriptions your team keeps renewing without checking

[9inches @0xnineinch]

@Klonzu i like this motion video

[9inches @0xnineinch]

@0xGenAi it's xmas gift lol

[9inches @0xnineinch]

@skunzu rea is ultimative tool for such tasks

[9inches @0xnineinch]

Haggle Bot + Jev is a ruthless 24/7 AI team that could cut $100K+ from your software subscriptions and vendor bills

Elon Musk puts Haggle Bot’s potential at up to $6 trillion in annual savings if a 5% improvement in enterprise cost efficiency scales across the global economy

Jev handles the quick decisions: which bill needs investigation, which needs more data, and which can reuse a verified finding, so Haggle spends its time on expenses worth reviewing

connect your invoices, vendor contracts, and software usage records, then build a recurring workflow that catches wasted spend before the next renewal

if you’re running an agency, a software company, or a business juggling dozens of subscriptions, there’s a clear place to start

the full setup:

→ give Jev 5 possible next actions: reuse a verified result, request missing data, investigate a bill, prepare a vendor counter, or flag something for your review

→ let code compare each invoice with the previous version, checking prices, seat counts, renewal dates, and whether the same expense has already been investigated

→ send Haggle the contract, payment history, and actual usage, so it can look for unused seats, overlapping tools, unexpected price increases, and cheaper alternatives

→ have it turn each finding into a concrete proposal: which seats to remove, which subscription duplicates another tool, or what terms to ask the vendor to change

→ put cancellations and vendor messages in a review queue, with the supporting evidence, expected saving, and contract conditions attached

→ save every accepted finding with its inputs and outcome, so matching bills reuse the result while changed prices or usage trigger a fresh review

→ every morning, get one page showing new opportunities, decisions waiting on you, confirmed reductions, and what the AI workflow itself cost to run

run it for 7 days and compare the savings identified, the changes approved, and the money actually removed from future bills

keep the useful routines running around the clock, with a spending cap and clear rules for when the system needs you

save this and start with the subscriptions your team keeps renewing without checking

[9inches @0xnineinch]

https://t.co/qjC88AQUbw

[9inches @0xnineinch]

Jev + Opus 5.5 made my whole productivity stack 20x more efficient

6 commands, 1 vault, and no more paying Opus twice for the same answer — Jev routes the work, saved results handle repeats, and Opus tackles the hard stuff

the vault holds my notes, decisions, drafts, and unfinished work, with Jev selecting what each task needs and Opus working from that context instead of making me explain everything from scratch

/process → sorts my inbox and turns it into linked notes
/connect → surfaces connections across what I’ve already saved
/review → brings back unfinished tasks and decisions I forgot
/ask → selects relevant notes before Opus answers
/draft → turns those notes into posts without explaining everything again
/decide → routes routine choices through Jev and deeper work to Opus

the clever part: all 6 commands share the same vault, so a decision made during /review can guide the next /draft, and an answer from /ask stays available for future tasks instead of disappearing into another chat

before another paid call, the workflow checks whether there’s a usable saved result, lets Jev select the next action, and sends Opus the relevant evidence when deeper work is needed

the full Opus 5.5 + Jev setup below includes every prompt and shows how to connect the vault, routing, and saved results around your own recurring tasks ↓

[9inches @0xnineinch]

https://t.co/qjC88AQUbw

[9inches @0xnineinch]

@skunzu interesting combo fr

[9inches @0xnineinch]

@Synthexfr what's the total cost?

[9inches @0xnineinch]

@Klonzu good setup!

[9inches @0xnineinch]

@SofiaTachExpert saved this moment

[9inches @0xnineinch]

@barz3n he's one of the greates person in AI sphere

[9inches @0xnineinch]

@armouredme is this real??

[9inches @0xnineinch]

i genuinely don't understand why everyone is burning tokens explaining the same sh*t to AI every session

Boris Cherny, the creator of Claude Code, shared an idea that hit 22 million views: give your AI a second brain that remembers your codebase, learns from every fix, and stops making you pay for the same mistake twice

build that brain around Opus 5.5 + Jev, with files storing your project’s memory, Jev selecting the relevant notes, and Opus using them before starting another investigation

your stack, your preferences, yesterday’s decisions, and last month’s fixes become context the next session can actually use

the full setup:

→ create a memory folder in your project and point CLAUDE.md to it, keeping your coding preferences, architecture decisions, verified fixes, and unfinished work in separate notes

→ on day 1 you tell Opus “use pnpm, not npm” and have it save the rule, so on day 2 it reads that instruction before touching your dependencies

→ when you fix a bug, save what broke, what caused it, which files changed, and the check that proved the solution worked, giving future sessions a useful record of the fix

→ before each task, retrieve candidate notes and let Jev select what belongs in the context, so a login issue brings back authentication decisions instead of pages of unrelated design work

→ on day 30 a related task surfaces the bug you fixed three weeks ago, with Opus checking the saved solution against the current code before trying the same failed approaches again

→ correct it once and update the original rule, removing conflicting notes so the next session gets the lesson instead of making you repeat it

each completed task leaves something useful behind, and each new session has less work to rediscover

that can mean fewer repeated investigations, fewer unnecessary calls, and less money spent solving problems you already solved

start with one project and one recurring task, then compare the time and cost across sessions

my Opus 5.5 + Jev article below shows how to connect this project brain with routing, saved results, and targeted retries ↓

[9inches @0xnineinch]

https://t.co/qjC88AQUbw

[9inches @0xnineinch]

you can add it from Grok Bot marketplace:
https://t.co/pFtnwJpxBP

[9inches @0xnineinch]

holy f*ck, SpaceXAI’s own engineers just put 54 ready-made Grok Bots in your hands for $0

these 10 are the team’s recommended starting lineup, covering the work you keep putting off across writing, video editing, coding, research, and project management

built by Lauren Tan, Shub Gaur, Matt Palmer and more, with one clear job for each bot

01 dr eggbot → creates specialist Grok Bots around the work you want done

02 Import Bot → brings your existing AI workflows into Grok Bot

03 Projects Manager → coordinates projects and tasks for your specialist bots through Notion

04 Lingxi’s Engineer Bot → launches coding agents and monitors their pull requests

05 Startup QA Bot → checks your product and reports bugs with screenshots and reproduction steps

06 Video Editor → turns raw footage into short videos with captions, B-roll, and music

07 Stills & Clips Desk → extracts thumbnails, stills, and short clips from your footage

08 Writing Bot → drafts and revises your posts, emails, and documents

09 Cooper → delivers daily news briefings and meaningful competitor alerts in Slack

10 Haggle Bot → finds wasted SaaS spending and drafts vendor negotiations for review

open the marketplace, pick the job you hate doing, and hit Add, starting with something you already spend time on every week so you have a clear idea of what a useful result should look like

give that bot your context, connect the tools it needs, and hand it one concrete assignment with the files, examples, and requirements it needs to finish the work

the roles are already defined, so you can spend your time teaching it your workflow, reviewing the first result, and giving specific feedback that makes the next attempt more useful

i left the marketplace link in the comments so you can find these bots, check what each one does, and pick the first one for your own setup ↓

[monokern @monokern]

GROK BOT SPIED ON 8 COMPETITORS FOR 24 HOURS AND RECORDED ALL THEIR ONBOARDINGS ON VIDEO

by the evening one competitor's sales team had emailed it three times trying to book a demo call

it's Stalk Bot from the Grok Bot marketplace. it signs up for competitor products like a regular user and screen-records everything they show a new customer

→ first hour, 8 signups and the first onboarding saved, with a paywall on step 4
→ evening, every new email from that SDR comes with a bigger discount
→ night, another company tests two pricing pages and the bot catches both versions
→ next morning, a third one opens 5 enterprise sales roles in a single day

usually the only way to see a competitor's sales funnel from the inside is to sign up yourself, and then their team knows exactly who you are

their emails turned out to be more useful than their websites, because you see which features they push and how big a discount they offer once a lead goes quiet

8 onboardings on video, 23 emails from their sales teams, 2 versions of one pricing page, each hidden from half their visitors

and that SDR is still waiting for the bot to pick a time for the demo

[9inches @0xnineinch]

THIS IS A PURE GOLD

MIT students shared 9 use cases where Opus 5.5 + Jev makes your workflow 20x more efficient

Jev handles the quick decisions about what deserves more work while Opus develops the answers that need deeper reasoning, giving each part of the setup a clear job

route requests to the right model, sort incoming emails, and rank the documents worth reading before another expensive investigation begins

check suspicious inputs, decide whether a tool should run, and flag uncertain actions for human review before the agent moves forward

the same approach extends to labeling large datasets, scoring model outputs, and choosing actions in games, bots, or other live systems

each of these is a small decision you’re probably still paying a larger model to make, and the cost adds up when it happens across thousands of tasks

move those decisions into Jev and give Opus the selected work, making the projects you kept putting off cheaper to build and run

[9inches @0xnineinch]

https://t.co/qjC88AQUbw

[9inches @0xnineinch]

STOP SPENDING 20x MORE ON OPUS 5.5 THAN YOU NEED TO

1,200 incoming reports. 60 worth investigating. Jev filters the queue before every submission becomes an expensive Opus task

start in shadow mode. your workflow keeps running while Jev checks the same reports and logs what it would investigate, skip, or send for review

after a week, compare its choices with the actual work. see which calls you could have avoided and which important reports Jev would have missed

then let it route the tasks. Opus investigates the selected reports, while code tracks completed work and catches unchanged submissions

you’re cutting the queue before it reaches the expensive model

my article below covers the setup and how to measure whether that filter earns its place ↓

[9inches @0xnineinch]

STOP SPENDING 20x MORE ON OPUS 5.5 THAN YOU NEED TO

1,200 incoming reports. 60 worth investigating. Jev filters the queue before every submission becomes an expensive Opus task

start in shadow mode. your workflow keeps running while Jev checks the same reports and logs what it would investigate, skip, or send for review

after a week, compare its choices with the actual work. see which calls you could have avoided and which important reports Jev would have missed

then let it route the tasks. Opus investigates the selected reports, while code tracks completed work and catches unchanged submissions

you’re cutting the queue before it reaches the expensive model

my article below covers the setup and how to measure whether that filter earns its place ↓

[9inches @0xnineinch]

https://t.co/qjC88AQUbw

[9inches @0xnineinch]

HAIKU 5.5 JUST FOUND OUT WHY OPUS COSTS MORE.

Cloth physics from 4 to 4,096 faces.

Haiku 5.5 — 286.9K tokens · ~$3.08
Opus 5.5 — 194.0K tokens · ~$10.62

Same prompt. Both built from scratch.

but Opus 5.5 + Jev beats an all-Haiku workflow on both cost and output quality.

Jev handles focused decisions and batches checks. code skips repeated inputs. Opus gets selected tasks with clear requirements.

keep the stronger output and cut the calls you never needed.

my article below breaks down the setup and the math behind the 20x efficiency target ↓

[9inches @0xnineinch]

https://t.co/qjC88AQUbw