Everything in the atlas

It writes and it administers — ETL extraction and database user management are not read operations, and the same connection that profiles a column can move data out of the tenant. Masking is only as good as the configuration naming the fields, so an unlisted sensitive column is an unmasked one.

Its README says it plainly and so does this entry: this is not a simulation, every tool call reaches a live BW system, and write operations produce real changes. BW modelling objects are the definition of a warehouse — the blast radius of a wrong one is every report built on it.

It reaches a data warehouse, so what an agent can see is business data by definition — the SQL sanitisation and PII filtering in the server are there because of that, not instead of a decision about it. And the guidance is Datasphere-shaped: nothing here transfers to a HANA Cloud database used directly.

Ninety tools is a large surface to put in front of a model at once, and client-credentials OAuth means the tenant sees one identity rather than the person asking. Content transport and user management sit in the same tool list as reading a story.

When user-dependent responses are cached without user-aware or explicit operation keys, or translated responses share keys across locales. These cases require key isolation; locale-aware keys are off by default.

It writes and it administers — ETL extraction and database user management are not read operations, and the same connection that profiles a column can move data out of the tenant. Masking is only as good as the configuration naming the fields, so an unlisted sensitive column is an unmasked one.

It reaches a data warehouse, so what an agent can see is business data by definition — the SQL sanitisation and PII filtering in the server are there because of that, not instead of a decision about it. And the guidance is Datasphere-shaped: nothing here transfers to a HANA Cloud database used directly.

Where the release and programming model offer durable locks, as RAP draft handling does, the README states that those solve the problem at the kernel level and the lock manager is for where that is not an option. A running handler occupies one background work process for as long as any lock is held, and only generated `ENQUEUE_*` modules can be locked.

Not an end-user app: it reads from any table the user names and writes into tables of the Z/Y namespace, with no authorisation check of its own, so the README asks a project to add its own checks on `S_TABU_DIS`/`S_TABU_NAM` and to restrict who may run it before it is used beyond a development system. XLSX upload and download are not ready for ABAP Cloud yet, and of the upload apps only the JSON one writes the file's rows back to the database.

It writes and it administers — ETL extraction and database user management are not read operations, and the same connection that profiles a column can move data out of the tenant. Masking is only as good as the configuration naming the fields, so an unlisted sensitive column is an unmasked one.

Its README says it plainly and so does this entry: this is not a simulation, every tool call reaches a live BW system, and write operations produce real changes. BW modelling objects are the definition of a warehouse — the blast radius of a wrong one is every report built on it.

Ninety tools is a large surface to put in front of a model at once, and client-credentials OAuth means the tenant sees one identity rather than the person asking. Content transport and user management sit in the same tool list as reading a story.

Nothing about it is readable from outside the package: there is no repository, no issue tracker, and the npm page carries no readme, so the README, the examples and the changelog can be read only from the tarball or a CDN mirror, and a bug has nowhere to be filed. Provider URLs are called from the plugin with API keys taken from environment variables in a .env file, so on BTP no call resolves through a destination and the keys live in the environment. The npm line stops at 1.0.1 on 3 November 2025; the changelog shipped inside that version is headed v2.1, so the version npm shows and the version the author counts are not the same.

Not for a table whose contents matter, and not for one chosen at run time: the name of the table is fixed in the calling code, and the maintainer's own warning is "only hardcode the table name, don't use with important tables, use at own risk!"

The app writes to whatever table is passed in the URL parameter `table` and carries no authorisation check of its own, so past a development system it needs authorisation checks added and a restriction on who may run it.

The metadata is inconsistent: the declared repository is github.com/acme/search-vocab and the README installs @acme/search-vocab, neither of which matches the published name. One release, November 2025.

It changes what DELETE means for every client of the annotated projection, so anything that must actually remove a row — an erasure request, a retention job — has to go around it.