The car starts, but the idle hunts, or it sits slightly low and shakes, or it stumbles on the first touch of throttle and then runs fine. It was smooth before the intake went on. Above idle the car behaves normally, and that is the most useful clue in the whole complaint.
An idle-only fault is a metering fault at low airflow. At idle the throttle is nearly closed, airflow is small, and a fixed amount of error is a large proportion of it. The same leak that is harmless at a cruise is disruptive at idle, which is why the symptom concentrates there.
At small throttle openings the engine is most sensitive to unmetered air. A mass air flow system counts what passes the sensor, so anything entering downstream of the sensor is air the engine burns but the control unit does not know about. The mixture leans out, the idle speed rises, the control unit corrects, and the result is a hunt.
A speed density system has a different weakness. It infers airflow from manifold pressure, so a leak downstream of the throttle body shows as a pressure reading that does not match the throttle position, and the idle correction heads the wrong way. The mechanism differs, but the observable behaviour and the first place to look are the same.
Fitting an intake means disconnecting and reconnecting lines: breather hoses, purge lines, vacuum takeoffs and sometimes a resonance chamber. Any of them left off, pushed on at an angle, or refitted with a cracked end becomes an unmetered air path once the engine is running.
Check them by hand rather than by eye. Pull each line gently at its connection; one that slides off with no resistance was never seated. Look for a line stretched to reach a new position, because that stretch pulls the end off the barb as the engine moves on its mounts.
A mass air flow sensor is calibrated to the housing it sits in, and it is sensitive to how air approaches it. A new tube with a different bore, a different bend immediately ahead of the sensor, or the sensor rotated out of its original orientation will change the reported airflow at idle, where flow is slowest and least stable.
This is a scaling error rather than a leak, and it behaves differently: it is present whenever the engine runs, not only at idle, but it is most obvious at idle because that is where resolution is coarsest. It also tends to be consistent rather than intermittent, which is a useful distinction in itself.
Look at the long term fuel trim at idle and at a steady cruise. A leak produces a large correction at idle that shrinks toward cruise. A sensor scaling error produces a correction that stays roughly proportional to airflow across the whole range.
Short term behaviour confirms it. A leak makes the short term trim swing as the idle hunts. A scaling error makes it sit at a steady offset. Watching trims at idle and at a steady road speed separates the two families of fault faster than any amount of inspecting.
Refit the original intake as a test. If the idle returns immediately, the fault came with the part or with the installation, and you can separate those by refitting the new part carefully with fresh clamps and every line checked. If the idle stays rough with the original back on, something else changed at the same time.
Then check the throttle body and the idle relearn, because a battery disconnected during the job, or a throttle body disturbed, can leave the learned idle position wrong until the car relearns it. Where the intake itself is suspect, compare its bore and sensor position against the original; a cold air intake that keeps the metering section intact will not disturb the idle, and a correctly fitted air filter element is part of that same metering path.
