Added new PumpFun AMM buy event to stg_decoded_swaps#9428
Open
megavel wants to merge 18 commits intoduneanalytics:mainfrom
Open
Added new PumpFun AMM buy event to stg_decoded_swaps#9428megavel wants to merge 18 commits intoduneanalytics:mainfrom
megavel wants to merge 18 commits intoduneanalytics:mainfrom
Conversation
|
CLA Assistant Lite bot All contributors have signed the CLA ✍️ ✅ |
github-actions
bot
added
WIP
|
No actionable comments were generated in the recent review. 🎉 ℹ️ Recent review info⚙️ Run configurationConfiguration used: Organization UI Review profile: CHILL Plan: Pro Run ID: 📒 Files selected for processing (1)
📝 WalkthroughWalkthroughAdds a new incremental staging model decoding Pumpswap "new buy" instructions, an intermediate UNION ALL view consolidating decoded swaps, updates the backfill to consume the union and advances project start dates; plus schema and source entries and minor formatting fixes. Changes
Sequence Diagram(s)sequenceDiagram
participant InstructionCalls as InstructionCalls\n(solana.instruction_calls)
participant PumpAmm as PumpAmm\n(pump_amm_evt_buyevent)
participant StgOld as StgOld\n(pumpswap_solana_stg_decoded_swaps)
participant StgNew as StgNew\n(pumpswap_solana_stg_decoded_newevent)
participant IntAll as IntAll\n(pumpswap_solana_int_all_swaps)
participant Backfill as Backfill\n(pumpswap_solana_base_trades_backfill)
InstructionCalls->>PumpAmm: join/filter by discriminator & execution/tx_success
InstructionCalls->>StgNew: decode new buy events, compute surrogate_key,\nblock_month/block_date, dedupe (rn=1)
InstructionCalls->>StgOld: existing decode flow (unchanged)
StgOld->>IntAll: supply decoded swap rows
StgNew->>IntAll: supply decoded new-event rows
IntAll->>Backfill: backfill reads consolidated swaps
Estimated code review effort🎯 3 (Moderate) | ⏱️ ~20 minutes Possibly related PRs
Suggested reviewers
🚥 Pre-merge checks | ✅ 2 | ❌ 1❌ Failed checks (1 inconclusive)
✅ Passed checks (2 passed)
✏️ Tip: You can configure your own custom pre-merge checks in the settings. Comment |
Author
|
I have read the CLA Document and I hereby sign the CLA |
![coderabbitai[bot]](https://avatars.githubusercontent.com/in/347564?s=60&v=4){kind=small}
There was a problem hiding this comment.
Actionable comments posted: 1
🤖 Prompt for all review comments with AI agents
Verify each finding against the current code and only fix it if needed.
Inline comments:
In
`@dbt_subprojects/solana/models/_sector/dex/pumpswap/staging/pumpswap_solana_stg_decoded_swaps.sql`:
- Around line 95-113: The join between g and f is currently only on tx_id
causing N:N matches; change it to a 1:1 instruction-scoped join (add the
instruction-level key such as instruction_index or instruction_id: e.g., ON
g.tx_id = f.tx_id AND g.instruction_index = f.instruction_index) and include
block_date equality (AND g.block_date = f.block_date) so the relationship is
unique and partition-prunable; also switch any filters using block_time to use
block_date for partition pruning (replace incremental_predicate('f.block_time')
and incremental_predicate('g.block_time') with
incremental_predicate('f.block_date') and incremental_predicate('g.block_date')
and ensure WHERE includes AND g.block_date >= DATE '{{ project_start_date }}' /
f.block_date counterpart).
ℹ️ Review info
⚙️ Run configuration
Configuration used: Organization UI
Review profile: CHILL
Plan: Pro
Run ID: e7f67dfb-fb8d-46b4-8d85-a91942a3cef3
📒 Files selected for processing (2)
dbt_subprojects/solana/models/_sector/dex/pumpswap/pumpswap_solana_pools.sqldbt_subprojects/solana/models/_sector/dex/pumpswap/staging/pumpswap_solana_stg_decoded_swaps.sql
Comment on lines
+95
to
+113
| JOIN {{ source('solana', 'instruction_calls') }} f | ||
| ON g.tx_id = f.tx_id | ||
| AND bytearray_substring(f.data, 1, 8) = 0xc62e1552b4d9e870 | ||
| AND f.executing_account = 'pAMMBay6oceH9fJKBRHGP5D4bD4sWpmSwMn52FMfXEA' | ||
| AND f.tx_success = true | ||
| {% if is_incremental() %} | ||
| AND {{ incremental_predicate('f.block_time') }} | ||
| {% else %} | ||
| AND f.block_time >= TIMESTAMP '{{ project_start_date }}' | ||
| {% endif %} | ||
| WHERE bytearray_substring(g.data, 1, 16) = 0xe445a52e51cb9a1d67f4521f2cf57777 | ||
| AND g.executing_account = 'pAMMBay6oceH9fJKBRHGP5D4bD4sWpmSwMn52FMfXEA' | ||
| AND g.is_inner = true | ||
| AND g.tx_success = true | ||
| AND length(g.data) = 447 | ||
| {% if is_incremental() %} | ||
| AND {{ incremental_predicate('g.block_time') }} | ||
| {% else %} | ||
| AND g.block_time >= TIMESTAMP '{{ project_start_date }}' |
There was a problem hiding this comment.
Tighten the g↔f join to an instruction-scoped key.
Right now this branch matches g and f on tx_id only. If a transaction contains more than one qualifying PumpFun instruction, every matching f row can pair with every matching g row, which will duplicate swaps and attach the wrong account_pool_*_token_account values. Please join on the smallest instruction-level key that makes the relationship 1:1 for this tx, and include block_date in the join/filter so the raw-table self-join stays prunable.
As per coding guidelines, "Filter on block_date (not block_time) for partition pruning, and always include partition columns in WHERE and JOIN conditions".
🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed.
In
`@dbt_subprojects/solana/models/_sector/dex/pumpswap/staging/pumpswap_solana_stg_decoded_swaps.sql`
around lines 95 - 113, The join between g and f is currently only on tx_id
causing N:N matches; change it to a 1:1 instruction-scoped join (add the
instruction-level key such as instruction_index or instruction_id: e.g., ON
g.tx_id = f.tx_id AND g.instruction_index = f.instruction_index) and include
block_date equality (AND g.block_date = f.block_date) so the relationship is
unique and partition-prunable; also switch any filters using block_time to use
block_date for partition pruning (replace incremental_predicate('f.block_time')
and incremental_predicate('g.block_time') with
incremental_predicate('f.block_date') and incremental_predicate('g.block_date')
and ensure WHERE includes AND g.block_date >= DATE '{{ project_start_date }}' /
f.block_date counterpart).
- Create separate staging model for new buy event (0xe445a52e51cb9a1d) - Implement 1:1 join using outer_instruction_index for deduplication - Add ROW_NUMBER() to handle potential duplicate rows from join - Improve partition pruning with block_date filters - Create intermediate model to union old and new event staging models - Update schema.yml with new model documentation
This comment was marked as duplicate.
This comment was marked as duplicate.
Sorry, something went wrong.
github-actions
bot
added
ready-for-review
There was a problem hiding this comment.
Actionable comments posted: 1
🧹 Nitpick comments (2)
dbt_subprojects/solana/models/_sector/dex/pumpswap/staging/pumpswap_solana_stg_decoded_newevent.sql (1)
40-40: Use explicitINNER JOINinstead of bareJOIN.Per coding guidelines, explicit join types should be used for clarity.
♻️ Suggested fix
- JOIN {{ source('solana', 'instruction_calls') }} f + INNER JOIN {{ source('solana', 'instruction_calls') }} f🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@dbt_subprojects/solana/models/_sector/dex/pumpswap/staging/pumpswap_solana_stg_decoded_newevent.sql` at line 40, The query uses a bare "JOIN" which needs to be made explicit; replace the occurrence of "JOIN {{ source('solana', 'instruction_calls') }} f" with "INNER JOIN {{ source('solana', 'instruction_calls') }} f" to follow the coding guideline requiring explicit join types (look for the token instruction_calls and alias f in the pumpswap_solana_stg_decoded_newevent SQL to locate the exact line).dbt_subprojects/solana/models/_sector/dex/pumpswap/intermediate/pumpswap_solana_int_all_swaps.sql (1)
9-17: AvoidSELECT *- specify exact columns needed.Using
SELECT *on these staging tables violates the columnar storage guideline and can degrade query performance. Enumerate the required columns explicitly to benefit from Dune's columnar storage optimization. As per coding guidelines: "Never useSELECT *on large tables; specify exact columns needed due to Dune's columnar storage".♻️ Suggested refactor
SELECT - * -FROM {{ ref('pumpswap_solana_stg_decoded_swaps')}} + block_slot + , block_month + , block_date + , block_time + , inner_instruction_index + , swap_inner_index + , outer_instruction_index + , outer_executing_account + , tx_id + , tx_index + , pool + , user_account + , account_user_base_token_account + , account_user_quote_token_account + , account_pool_base_token_account + , account_pool_quote_token_account + , account_protocol_fee_recipient_token_account + , base_amount + , is_buy + , surrogate_key +FROM {{ ref('pumpswap_solana_stg_decoded_swaps') }} UNION ALL SELECT -* -FROM {{ref('pumpswap_solana_stg_decoded_newevent')}} + block_slot + , block_month + , block_date + , block_time + , inner_instruction_index + , swap_inner_index + , outer_instruction_index + , outer_executing_account + , tx_id + , tx_index + , pool + , user_account + , account_user_base_token_account + , account_user_quote_token_account + , account_pool_base_token_account + , account_pool_quote_token_account + , account_protocol_fee_recipient_token_account + , base_amount + , is_buy + , surrogate_key +FROM {{ ref('pumpswap_solana_stg_decoded_newevent') }}🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@dbt_subprojects/solana/models/_sector/dex/pumpswap/intermediate/pumpswap_solana_int_all_swaps.sql` around lines 9 - 17, Replace the two "SELECT *" uses with an explicit, matching column list: inspect the staging models referenced by ref('pumpswap_solana_stg_decoded_swaps') and ref('pumpswap_solana_stg_decoded_newevent'), enumerate the exact columns you need (e.g., tx_hash, block_number, timestamp, sender, recipient, amount, token_address, etc.), and use that same ordered column list in both SELECTs so UNION ALL aligns. Ensure you only include needed columns to respect Dune's columnar-storage guideline and update any downstream references that expect specific columns from this intermediate model.
🤖 Prompt for all review comments with AI agents
Verify each finding against the current code and only fix it if needed.
Inline comments:
In
`@dbt_subprojects/solana/models/_sector/dex/pumpswap/staging/pumpswap_solana_stg_decoded_newevent.sql`:
- Line 9: The config currently sets incremental_predicates =
[incremental_predicate('DBT_INTERNAL_DEST.block_date')] but the model and its
WHERE filters use block_time; update the config to use incremental_predicates =
[incremental_predicate('DBT_INTERNAL_DEST.block_time')] so the incremental
predicate matches the query filters (see incremental_predicate usages in the
body) and ensures correct time-series incremental behavior.
---
Nitpick comments:
In
`@dbt_subprojects/solana/models/_sector/dex/pumpswap/intermediate/pumpswap_solana_int_all_swaps.sql`:
- Around line 9-17: Replace the two "SELECT *" uses with an explicit, matching
column list: inspect the staging models referenced by
ref('pumpswap_solana_stg_decoded_swaps') and
ref('pumpswap_solana_stg_decoded_newevent'), enumerate the exact columns you
need (e.g., tx_hash, block_number, timestamp, sender, recipient, amount,
token_address, etc.), and use that same ordered column list in both SELECTs so
UNION ALL aligns. Ensure you only include needed columns to respect Dune's
columnar-storage guideline and update any downstream references that expect
specific columns from this intermediate model.
In
`@dbt_subprojects/solana/models/_sector/dex/pumpswap/staging/pumpswap_solana_stg_decoded_newevent.sql`:
- Line 40: The query uses a bare "JOIN" which needs to be made explicit; replace
the occurrence of "JOIN {{ source('solana', 'instruction_calls') }} f" with
"INNER JOIN {{ source('solana', 'instruction_calls') }} f" to follow the coding
guideline requiring explicit join types (look for the token instruction_calls
and alias f in the pumpswap_solana_stg_decoded_newevent SQL to locate the exact
line).
ℹ️ Review info
⚙️ Run configuration
Configuration used: Organization UI
Review profile: CHILL
Plan: Pro
Run ID: b06fce8d-c793-4a59-8088-52c468184966
📒 Files selected for processing (5)
dbt_subprojects/solana/models/_sector/dex/pumpswap/_schema.ymldbt_subprojects/solana/models/_sector/dex/pumpswap/intermediate/pumpswap_solana_int_all_swaps.sqldbt_subprojects/solana/models/_sector/dex/pumpswap/pumpswap_solana_base_trades_backfill.sqldbt_subprojects/solana/models/_sector/dex/pumpswap/staging/pumpswap_solana_stg_decoded_newevent.sqldbt_subprojects/solana/models/_sector/dex/pumpswap/staging/pumpswap_solana_stg_decoded_swaps.sql
✅ Files skipped from review due to trivial changes (1)
- dbt_subprojects/solana/models/_sector/dex/pumpswap/staging/pumpswap_solana_stg_decoded_swaps.sql
...projects/solana/models/_sector/dex/pumpswap/staging/pumpswap_solana_stg_decoded_newevent.sql
Show resolved
Hide resolved
- Added explicit column list to int_all_swaps UNION ALL branches - Change JOIN to explicit INNER JOIN in stg_decoded_newevent - Switch new buy event from raw instruction_calls to decoded source table - Changed project start date to 2026-03-06 for testing only
…book into pumpswap-new-event Merge remote changes.
There was a problem hiding this comment.
Actionable comments posted: 2
🤖 Prompt for all review comments with AI agents
Verify each finding against the current code and only fix it if needed.
Inline comments:
In
`@dbt_subprojects/solana/models/_sector/dex/pumpswap/pumpswap_solana_base_trades_backfill.sql`:
- Line 19: The fallback date assigned to the Jinja variable begin ('{% set begin
= ... %}') is out of sync with the model config; update the fallback value used
when model.batch is absent so it matches the model config(begin = '2025-02-20')
— i.e., change the begin fallback to '2025-02-20' in the
pumpswap_solana_base_trades_backfill.sql template so the query backfills from
the same date as the model config.
In
`@dbt_subprojects/solana/models/_sector/dex/pumpswap/staging/pumpswap_solana_stg_decoded_swaps.sql`:
- Line 14: The project_start_date was narrowed to '2026-03-06', which will drop
legacy swaps (2025-02-20 through 2026-03-05); revert project_start_date in
pumpswap_solana_stg_decoded_swaps (the project_start_date variable) to the
previous/original value that covers 2025-02-20 (or set it to an earlier
universal start) so that pumpswap_solana_int_all_swaps can reconstruct full
history, or remove the artificial restriction entirely so legacy swaps are
preserved.
ℹ️ Review info
⚙️ Run configuration
Configuration used: Organization UI
Review profile: CHILL
Plan: Pro
Run ID: be24c4f3-6d4a-4206-ad3f-62d0bdce40ae
📒 Files selected for processing (6)
dbt_subprojects/solana/models/_sector/dex/pumpswap/_schema.ymldbt_subprojects/solana/models/_sector/dex/pumpswap/intermediate/pumpswap_solana_int_all_swaps.sqldbt_subprojects/solana/models/_sector/dex/pumpswap/pumpswap_solana_base_trades_backfill.sqldbt_subprojects/solana/models/_sector/dex/pumpswap/staging/pumpswap_solana_stg_decoded_newevent.sqldbt_subprojects/solana/models/_sector/dex/pumpswap/staging/pumpswap_solana_stg_decoded_swaps.sqldbt_subprojects/solana/models/_sources.yml
🚧 Files skipped from review as they are similar to previous changes (2)
- dbt_subprojects/solana/models/_sector/dex/pumpswap/staging/pumpswap_solana_stg_decoded_newevent.sql
- dbt_subprojects/solana/models/_sector/dex/pumpswap/intermediate/pumpswap_solana_int_all_swaps.sql
| }} | ||
|
|
||
| {% set begin = '2025-02-20' %} | ||
| {% set begin = '2026-03-06' %} |
There was a problem hiding this comment.
Keep the SQL fallback begin aligned with the model config.
This no longer matches Line 11's config(begin = '2025-02-20'). When model.batch is absent, the query now skips the first year of Pumpswap trades even though the microbatch config still says to backfill from February 20, 2025.
Suggested fix
-{% set begin = '2026-03-06' %}
+{% set begin = '2025-02-20' %}📝 Committable suggestion
‼️ IMPORTANT
Carefully review the code before committing. Ensure that it accurately replaces the highlighted code, contains no missing lines, and has no issues with indentation. Thoroughly test & benchmark the code to ensure it meets the requirements.
Suggested change
| {% set begin = '2026-03-06' %} | |
| {% set begin = '2025-02-20' %} |
🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed.
In
`@dbt_subprojects/solana/models/_sector/dex/pumpswap/pumpswap_solana_base_trades_backfill.sql`
at line 19, The fallback date assigned to the Jinja variable begin ('{% set
begin = ... %}') is out of sync with the model config; update the fallback value
used when model.batch is absent so it matches the model config(begin =
'2025-02-20') — i.e., change the begin fallback to '2025-02-20' in the
pumpswap_solana_base_trades_backfill.sql template so the query backfills from
the same date as the model config.
| }} | ||
|
|
||
| {% set project_start_date = '2025-02-20' %} | ||
| {% set project_start_date = '2026-03-06' %} |
There was a problem hiding this comment.
Don't narrow the legacy swaps model to March 2026.
This staging model still carries the pre-new-event buy/sell history. Moving project_start_date to 2026-03-06 means any full refresh or first-time build will miss swaps from 2025-02-20 through 2026-03-05, and pumpswap_solana_int_all_swaps cannot reconstruct them because the new-event model only covers the new discriminator path.
Suggested fix
-{% set project_start_date = '2026-03-06' %}
+{% set project_start_date = '2025-02-20' %}🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed.
In
`@dbt_subprojects/solana/models/_sector/dex/pumpswap/staging/pumpswap_solana_stg_decoded_swaps.sql`
at line 14, The project_start_date was narrowed to '2026-03-06', which will drop
legacy swaps (2025-02-20 through 2026-03-05); revert project_start_date in
pumpswap_solana_stg_decoded_swaps (the project_start_date variable) to the
previous/original value that covers 2025-02-20 (or set it to an earlier
universal start) so that pumpswap_solana_int_all_swaps can reconstruct full
history, or remove the artificial restriction entirely so legacy swaps are
preserved.
…book into pumpswap-new-event
…wap_with_transfers - Split swaps_with_transfers into UNION of old and new events - Old events: use INNER JOIN with transfers (preserves existing behavior) - New events: bypass transfer join, use decoded quote_token_amount directly
…book into pumpswap-new-event
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Thank you for contributing to Spellbook 🪄
Please open the PR in draft and mark as ready when you want to request a review.
Description:
Added support for a new PumpFun AMM buy event type that was previously not part of pumpdotfun-solana schema.
Changes
Added third UNION topumpswap_solana_stg_decoded_swaps.sqlto include new buy eventDecodes instruction with discriminator0xe445a52e51cb9a1d67f4521f2cf57777directly fromsolana.instruction_calls0xc62e1552b4d9e870) to extract pool token account addresses fromaccount_argumentsneeded the new eventDetails
New event is 447 bytes different from the first buy event which was 432 byte in size with specific byte-offset mapping-~~ Executes on
pAMMBay6oceH9fJKBRHGP5D4bD4sWpmSwMn52FMfXEA(PumpFun AMM)~~Follows same schema as existing buy/sell events for consistencyNew Changes
Architecture:
stg_decoded_newevents.sql- separate staging model for new event decodingint_all_swaps.sql- intermediate model that unions existing and new event staging modelsint_all_swapsinstead ofstg_decoded_swapsTechnical Improvements:
outer_instruction_indexandblock_dateon newevent model to avoid N:N matches recomended by the botROW_NUMBER()deduplication to handle potential duplicate rows from joinFiles Changed
models/pumpswap/solana/staging/stg_decoded_newevents.sql(new)models/pumpswap/solana/intermediate/int_all_swaps.sql(new)latest Modified Files:
The new events have a
quote_amountcolumn built into the decoded source, while old events need to fetch it from the transfers table.models/_sector/dex/pumpswap/staging/pumpswap_solana_stg_decoded_swaps.sqlquote_amountcolumn (NULL)models/_sector/dex/pumpswap/pumpswap_solana_base_trades_backfill.sqlstg_decoded_swapstoint_all_swapsswapsCTE to includequote_amountThis way:
Old events have quote_amount = NULL and gets the quote_amount from transfers source
New events have quote_amount populated and skip the transfer join
Validated the modified models with recent data (2026-03-06+) instead of (2025-02-20) just for testing purpose which will be changed
Here's also a reference dune query showing the entire decoded instruction data and txns for the new pump fun: AMM buy event:
https://dune.com/queries/6777900/