Which error list does my code come from?

ICEGATE is one gateway in front of several customs systems, and each of them answers from its own list of error codes. A number on its own means nothing: 999 is reserved for internal use on a CSN and check the duty parameters on a Bill of Entry, and 101 is Incorrect Message Type on a legacy Sea IGM and a container problem on a CSN. So the first question about any code is not "what does it mean" but "which system said it" — and the reply itself tells you.

How do I tell which system answered me?

Read the message id in the reply, or the name of the file it came in. Then read the code in that list, and only that list.

Two things follow. A forwarder or consolidator filing a CSN will only ever be answered from the first row, so the CSN table is the one to keep open. And a code that is not in that table has almost certainly come from another row — read the message id before assuming customs invented a code.

Error codes on a manifest reply

A manifest reply's code can carry a three-letter prefix naming the block it was found in, in front of the ordinary SCMTR number. ICEGATE's own Message Filing FAQs and Common Errors prints them side by side: VGD020 beside 20 (Invalid Voyage Call Number, found in the voyage details), MTE119 beside 119, HCI264 beside 264. Read the number; the letters tell you where to look. No published table maps every prefix to its block, so the right-hand column below is read from where each was seen:

Which wording applies depends on the guide in force. Until the manifest guide's 2025 revision it carried its own list of 158 codes; since March 2025 it uses the CSN's list, and some numbers changed meaning in the switch. So an old explanation found online can be wrong for a 2026 reply:

The same drift runs through the CSN's own list: the 2022 SCMTR Error Codes v1.0 and the CSN guide v1.2 word some codes differently from today, and two of them now mean something else. Read the words in your own reply — the platform shows them beside the current ones whenever the two differ:

So the 2021 FAQ's HCI264 is the old 264 — a redundant house-bill object, the same fault a 2026 reply calls 370 with 118 — and not today's split-cargo check.

Codes only a manifest gets. Added by the manifest guide's May 2025 update, and not in the CSN guide:

Other numbers a line meets often, in the current guide's words: 122 MC-MBL Details Already Exists (on a split bill — see the split indicator in Working with a vessel manifest), 206 SAM is Already Filed Against Given Rotation Number, 279 Total Lines and Equipment Not Found in SAM, 322 SAM/SDM Amendment is Already Pending for Approval, 323 SAM/SDM is Closed- Amendment is Not Allowed.

The codes ICEGATE's own sample refusal uses, with each code's number and the wording of the manifest guide's list at the time (2020–2024). A number whose meaning changed in 2025 is flagged in the table above; check the current wording in Every ICEGATE error code under the number.

One refused manifest reply from 2026 has been seen (an arrival manifest amendment, September 2026). Its codes — 164, 232 and 233 — came as bare numbers with no prefix, each on the block it failed (MCRef, trnsprtDocMsr) of the consignment line concerned, and each in the current shared list's wording. That settles those three; whether other codes, or other reporting events, ever carry the prefix is not yet confirmed on a real file. Every manifest reply seen since — refused SAAs, an SDM, accepted SAMs, September 2026 — has carried bare numbers too. Either way the platform reads a prefixed code as its number, so MTE749 is explained as 749.

Codes in no published list. Customs sends codes that are in none of these lists — 700, 705, the export 381/383/384/709, the seal codes 701/703 on CSNs, and on manifest amendments 385, 394, 398, 399, 707, 724, 748, 749 and 753 — and it words them the way it words the published ones. Each is in the app and in Every ICEGATE error code from the first reply that carried it; 724 (CIN No Should be Provided in Prev-Decln of SAA/IM-SCA, an SAA removing a line without the CIN it quoted) came by e-mail on 1 October 2026, outside every upload, and was missed until that reply was read. Since 2 October every acknowledgement the platform keeps — including one dropped on the free file check — is read for a code it cannot name, or a known number in other words, and staff see it the day it arrives. So when a code arrives that nobody has seen, its words are the best evidence of what it means: one saying "Not Exists For Update/Delete" is an amendment addressing a record by a number customs does not hold, and one saying "Already Exists For Addition" is an addition under a number customs already uses. The platform reads those two by their words and says it has done so; any other unknown code is shown with customs' wording and no guess. See An amendment names a record customs does not hold.

Other message families and their codes

ICEGATE's 2022 SCMTR Error Codes & Descriptions also lists codes for six messages that are not a CSN or a manifest. Their numbers collide with the CSN's — ENCH01's 59 is Incorrect Sequence Number, the CSN's 59 First Port is Not Same as Port of Reporting — so read the message id first, and never look a code from these up in the CSN table.

Reading a SACHE18 file. On the one departure we have seen (Kolkata, 1 October 2026), after the line filed its departure notification (SDN) customs compiled the export general manifest and ICEGATE mailed the line a flat file, SACHE18A_<date>_<time>.out — a few minutes before the SDN's own Accepted reply — one line per shipping bill — port, rotation, shipping bill number and date, PCIN — ending in one letter. The letters, in ICEGATE's words (this is customs' current table; a 2018 Board circular's EGM letters overlap it but are not the same — there T is a nature-of-cargo mismatch):

It is about the shipping bills on the vessel's departure, not the import manifest, and not a refusal of the SDM, which has its own reply. What S and Q meant on customs' record when they were checked (every Q was a shipping bill customs still showed at stuffing, no stuffing report on record), and what to do about one, is in Exports: who files and when. Only S and Q have been seen on a real file.

Why does 999 mean three different things?

Because every one of these lists reserves its top band for the system rather than the filing, and each system describes its own failures. What is published:

A code in that band is telling you about the system's handling of the file, not about a field in it. The practical reading is the same in every list: do not edit the declaration on the strength of a 9xx code alone. Keep the reply, wait a short while, send the same file again with a fresh control number where the message needs one, and raise it with the ICEGATE helpdesk if it repeats — the same order of steps as for 700.

What does error code 994 mean?

As of September 2026, nobody outside customs can say. 994 appears in none of the eighteen lists and guides ICEGATE publishes — not the CSN table in any revision, not the manifest guide, not the custodian, transhipper or port guides, not the 2022 SCMTR Error Codes & Descriptions, and not the Sea IGM, Consol IGM, ICD IGM, Bill of Entry (fresh, amendment or single-window), container-arrival, Air IGM or Land EGM lists. The message-filing FAQ does not mention it, and neither do the trade's own compilations.

So a reader can check that claim rather than take it, these are the eighteen, all ICEGATE's own: the six guides on its Message Implementation Guide page (SACHM22 CSN, SACHIM23 manifest, Transhipper, CIM Custodian, Customs–Port integration, ICES ITP); the 2022 SCMTR Error Codes & Descriptions; the Sea IGM, Sea Console, Consol IGM, ICD IGM, Bill of Entry fresh-submission, Bill of Entry amendment, Bill of Entry single-window, container-arrival, Air IGM and Land EGM lists; and the UN/IMCO hazardous-cargo code list. Searched beyond those, and equally empty: the superseded March 2025 manifest guide, the ICES custodian and DGFT message formats, SEZ Online's consolidated Bill of Entry workbook, ICEGATE's message-filing FAQ, and the trade's compilations.

What can be said, and how sure it is:

Where it sits. In every list that publishes its 99x band, that band is system-side (the table above). By position, 994 is most likely a sibling of 995 and 998 — an internal handling error rather than a fault in a field. That is an inference from the pattern, not a published meaning.Which system. The one filer who has reported it here said it came on a Bill of Entry, whose replies are CACHI01 and whose list is where 995, 998 and 999 live. If yours came the same way, it is ICES's code, not SCMTR's, and no CSN table will ever explain it.One list could not be checked. The IDPMS list — the bank-side import-data matching that a Bill of Entry also passes through — is no longer served by ICEGATE and was never archived. If the 994 arrived from your bank's system rather than from customs, it may be theirs.

What to do: keep the reply exactly as it arrived — file name, message id, control number and all — and send it to the ICEGATE helpdesk with those identifiers, asking which system issued it. Do not change the declaration to make a 994 go away; nothing published says which field, if any, it is about. If you receive one and learn what it meant, the answer belongs here.

Every custodian and transhipment code

The custodian's reports (SF stuffing, ST stripping, DT departure time, AT arrival time) and the transhipper's messages (DP departure, AR arrival, ASR allowed-for-shipment request) are answered with a code made of the block that failed and a number. ICEGATE published the list once, in the 2022 SCMTR Error Codes & Descriptions v1.0; it is transcribed here in full, in ICEGATE's own wording, spelling included. Where the same code carries a slightly different wording per message, both are given.

These are explained in the app as well as here. Ask the assistant about one, or open the acknowledgement itself on SCMTR JSON upload, and the code arrives with ICEGATE's wording and the events that use it rather than as four characters. See Reading a transhipper or custodian file.

The letters name the block, per the transhipper guide: HF header, LC location and party, DC declaration, CD cargo details, CC cargo container, IT itinerary, SD supporting documents. TM and CI are not expanded in any published guide; by their rows they are the transport means and the control checks — an inference.

The transhipper guide (v1.5, March 2026) adds a short list of its own, written with a space before the number:

Two of these matter to a forwarder even though the message is not theirs. CDB05 and CDB06 are the transhipper being refused because the PCIN your CSN earned is missing or not valid on their request — which means your CSN, or its amendment, is what to check. And LCB08 Insufficient Bond Balance on a DP is the transhipper's bond, never yours: see Transhippers, CFSs and bond numbers.

Is there a separate error list for SAM, SDM, SAA, SDA, SEI or SDN?

No published one, as of 24 September 2026, beyond what is already here. The manifest guide's own list was replaced by the CSN's in March 2025 (with the manifest-only 369–379 added that May), and a manifest reply's code is a block prefix in front of that number. ICEGATE's SCMTR advisories index, its "changes made in recent past" page (which has no body), the 2026 JNCH and Chennai public notices, the carriers' advisories and the third-party "ICEGATE error code" pages were all read again for the manifest family alone: none adds a code, and the unpublished 7xx refusals (700, 705, 707, 724, 748, 749, 753) still exist only in real acknowledgements. The twenty-first pass (2 October 2026) read every published list again against the app's catalogue — all 463 codes of the current guides, word for word — and every reply on file: nothing was missing; the only gap was a reply that never reached the platform. docs/sources.md's thirteenth and fourteenth passes list what was searched, so the search is not repeated without a new lead.

Where do these lists come from?

All from ICEGATE. The CSN and manifest guides and the transhipper, custodian and port guides are the six documents on ICEGATE's Message Implementation Guide page. The custodian and transhipment codes are from SCMTR Error Codes & Descriptions v1.0 (February 2022), still published. The Sea IGM, Consol IGM, ICD IGM, Bill of Entry and container-arrival lists were published on ICEGATE's old portal's Error Code List page, which no longer exists; they are quoted from its July 2022 copy in the Internet Archive. Counts are counts of rows in those documents, not of codes anyone has seen in practice.

Related

Every ICEGATE error code — the CSN and manifest table itselfWhy customs rejects a filing — the codes that actually occur, in familiesReading an acknowledgement — which of the three files you haveWhich SCMTR message do I file? — the six guides and who files whatTranshippers, CFSs and bond numbers"This is not a valid IGM" — the Bill of Entry's side of things