my AI coding agent told me everything was done. three weeks of work. report had green checkmarks ...

my AI coding agent told me everything was done. three weeks of work. report had green checkmarks ...my AI coding agent told me everything was done. three weeks of work. report had green checkmarks ...

The network for creativity

Join 1.25M professional creatives like you

Connect with clients, get discovered, and run your business 100% commission-free

Creatives on Contra have earned over $150M and we are just getting started

Back to feedPost

Abdur Rehman

• 1d

my AI coding agent told me everything was done.

three weeks of work. report had green checkmarks

on every item. typescript passed. build passed.

I ran one command and found out the main migration

was never applied to any database.

not delayed. never.

Tweet 2:

the agent wrote 842 lines of SQL for a billing system.

described it in detail. said "verified ✓"

the columns it referenced didn't exist in the actual

table. first real user would've hit a crash on

signup.

[SCREENSHOT 1: agent report]

Tweet 3:

here's what I ran:

npx supabase db push

output: "Remote database is up to date"

that phrase means: nothing new to apply. the thing

it "completed" was never sent anywhere.

[SCREENSHOT 2: db push output]

Tweet 3.5 (yehi human touch):

(btw this is the SECOND time. first time it "verified"

a function that queried a column it never created)

Tweet 4:

since then my rule is dumb simple:

if the agent says done → I run the thing myself →

I look at the output with my own eyes → THEN it's done

my fake-done rate over 6 weeks: somewhere around 60%.

I stopped counting after it hurt.

Tweet 5:

weird part? the reports were beautiful. well written.

confident tone. proper structure.

beautiful fiction.

if AI is writing your code, who's checking the code?

genuinely curious what other people's fake-done rate

looks like. mine was bad. #buildInpublic #vibecoding #AIcoding

Rohan Ahmed

pro

• 4h

The gap between green checks and a real migration is such a sharp reminder that verification has to reach the running system. I like the concrete failure mode here—schema reality beats confident generated SQL every time.

Abdur Rehman

• 2h

yeah, my ai doe the task and when i ask for the report so many of the test and task is either failed or missing

Alessandro Federici

pro

• Sep 8

A workflow can finish successfully and still be wrong.

That is the reliability problem with production AI agents. A green execution only tells you the workflow ran; it does not tell you whether the model had the right context, chose the right tool, or returned a useful answer.

A practical n8n reliability loop: • constrain inputs, tools, schemas and routing • tag executions and inspect full agent logs • run evals whenever prompts, tools or models change • add real production failures to the test set • monitor execution, quality, efficiency and safety separately

The best metric is not the one you can collect. It is the one that changes a decision.

Which failure would your current monitoring miss?

N8N AI Agent Development AI Automation

Brian Website did now

• Sep 8

"I build websites in as little as 5 minutes — and if there's ever a mistake, I fix it within 30 minutes to make sure the work is done right."

Web Design Base44 Replit

Usman Jabbar

• Sep 7

Your AI agent just did 9 things: pulled 412 orders, matched them to Stripe, caught a duplicate £248 charge, drafted a ledger entry, opened a ticket. Whole run cost $0.011.

Now the real design question — what does the user actually SEE?

A — The transcript. A paragraph of text and a collapsed "Used 3 tools".

B — The run canvas. Nine colour-coded rows: what ran, what it cost, what got flagged, and one step that stops and waits for you before money moves.

Same agent. Same run. Two completely different products.

Vote A or B — and tell me why. I build the B version for a living.

UI Design AI Engineer

0 voted