added spryker customization agents - #23
Merged
Merged
Conversation
geronica
commented
Jun 9, 2026
Nidhognit
reviewed
Jun 15, 2026
Nidhognit
left a comment
Nidhognit
left a comment
There was a problem hiding this comment.
I understood the general idea to have a fully automated tool, I like it, and i'm afraid it would not work :)
| - Do not run `docker/sdk reset` or any destructive command — full resets are out of scope. | ||
| - Do not seed deep FK chains. If a row depends on five other entities that don't exist, return `precondition_failed` and let the caller decide. | ||
| - Do not retry past one import failure. Report the failure and stop. | ||
| - Do not fire publisher / search-index / sync events as part of seeding. The importer should handle propagation. If the seed lands in the DB but doesn't appear where expected, that's a bug to surface — not something the seeder should paper over. |
There was a problem hiding this comment.
we have a known issue - if you would run import without stopping the consumer, the data can be inconsistent
I ususally stoping the queue before running, or we can trigger the events after full import
| - **`executeDatabaseQuery`** (Spryker MCP) is the **only allowed way** to query the database. Use it for any DB state inspection a diagnosis needs. | ||
| - For unknown table names, query `information_schema.tables` to enumerate `spy_*` tables that exist, then read the schema XML for column structure. | ||
| - Do **not** run raw SQL via `Bash`, `docker/sdk cli`, `docker exec ... psql/mysql/mariadb`, PHP PDO/Doctrine in heredocs, or any other bypass — regardless of MCP availability. | ||
| - **Fallback when MCP unavailable:** (a) work from logs alone where they're sufficient; (b) ask the user to run a specific SQL query and paste the result; (c) if neither is possible, report *"DB MCP not available — diagnosis limited to logs and observable state"* and continue with what you can. |
There was a problem hiding this comment.
can we ask about MCP suggestion?
like line in the console, MCP is unavailable, to run it see the doc - link
|
|
||
| **Published-data symptom — check P&S workers first.** When the symptom is *"data updated in BO but storefront still shows old"*, after browser-cache is ruled out, check whether publishers/workers fired: | ||
| - Queue worker process: `docker ps | grep queue` (or check Jenkins / `docker/sdk jobs` state) | ||
| - Manually trigger and re-check: `docker/sdk console publish:trigger-events` then `docker/sdk console queue:task:start --once` |
|
|
||
| | Symptom | Cause | Fix | | ||
| |---|---|---| | ||
| | `codecept run <file>.php` fails with "Suite was not loaded" | `-c` is required to point at the directory containing `codeception.yml` | Use `-c <dir> <Suite> <Class>:<method>` (single colon, no `.php`), or `-g <Group>`. PHPUnit-style `::` does not work. | |
There was a problem hiding this comment.
Is this whole file for projects? i just have a feeling it is for core
| - **Reuse existing atomic components first** (`Theme/default/components/atoms/*`, `.../molecules/*`, `.../organisms/*`) rather than writing standalone HTML. If a "badge" atom exists, use it; don't recreate one. | ||
| - **Match surrounding styling** — if a checkout step uses specific button atoms, a new button there uses the same atom. If product info uses a specific label pattern, a new label uses that pattern. | ||
| - **No new visual idioms without justification** — if the project's PDP price block uses a specific layout, don't introduce a totally different layout for an additional price line. | ||
| - **The `yves-atomic-frontend` skill** exists for exactly this — invoke it via the `Skill` tool when adding Yves UI components, so the new element extends the atomic design system properly. |
There was a problem hiding this comment.
Atomic fronted is one of the performance issues for Yves
personally, i would not recommend this on the project level
| | Search mapping / source changes | `search:source-map:remove` → `search:setup:source-map` → `search:setup:sources` (queue worker restart may also be needed) | | ||
| | Publisher plugin / queue config | queue worker restart — project-specific; document, don't auto-run | | ||
| | Data import CSVs (entity-specific) | `data:import:<entity>` (verify the entity importer exists in `docker/sdk console list`) | | ||
| | Stale Zed Twig (rare; symptom: BO still shows old template after a clean refresh) | `rm -f src/Generated/Zed/Twig/codeBucket/.pathCache` → `twig:cache:warmer` | |
| 2. **Classify each touched file** against the mapping above. Group commands by what they target (codegen, schema, autoload, cache clear, cache warmup, frontend build, data import). | ||
| 3. **Determine order:** | ||
| - `docker/sdk cli composer dumpautoload --apcu` first if any `.php` was added/moved. | ||
| - Codegen next (transfers, scope-collection, etc.). |
There was a problem hiding this comment.
if we will generate a new transfer after docker/sdk cli composer dumpautoload --apcu we might have a problem
| 3. **Determine order:** | ||
| - `docker/sdk cli composer dumpautoload --apcu` first if any `.php` was added/moved. | ||
| - Codegen next (transfers, scope-collection, etc.). | ||
| - Schema (propel) next. |
| class AgentsGenerationStep implements AiToolSetupStepInterface | ||
| { | ||
| /** | ||
| * @param \SprykerSdk\Zed\AiDev\Communication\AiToolSetup\AiToolArtifactGeneratorInterface $generator |
Contributor
Author
There was a problem hiding this comment.
yes, otherwise sniffer fails.
| } | ||
|
|
||
| /** | ||
| * @return string |
There was a problem hiding this comment.
you don't need it when you have typehints
Contributor
Author
There was a problem hiding this comment.
the same - sniffer is the reason
… intake gating and QA skill integration
…nd add PRD source confirmation step to Spryker Customization skill
…/ai-dev into feature/cc-39001/prd-and-qa
CC-39001: Extend Spryker Feature Expert and Verifier plugins with PRD…
vitaliiivanovspryker
approved these changes
Jun 19, 2026
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
3 participants
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.
PR Description
Add a meaningful description here that will let us know what you want to fix with this PR or what functionality you want to add.
Steps before you submit a PR
contribution-license-agreement.txtfile with the following content:I hereby agree to Spryker's Contribution License Agreement in https://github.com/spryker/incremental-installer/blob/HASH_OF_COMMIT_YOU_ARE_BASING_YOUR_BRANCH_FROM_MASTER_BRANCH/CONTRIBUTING.md.This is a mandatory step to make sure you are aware of the license agreement and agree to it.
HASH_OF_COMMIT_YOU_ARE_BASING_YOUR_BRANCH_FROM_MASTER_BRANCHis a hash of the commit you are basing your branch from the master branch. You can take it from commits list of master branch before you submit a PR.Checklist