/sap-check-abap

name: sap-check-abap
description: |
  Validates ABAP source (report / program / FM / class / include) before
  deployment — one skill, dimension-dispatched: naming (variable conventions),
  type (DDIC type validity), unused (dead variables), sql (SELECT/UPDATE/DELETE
  fields vs table defs), contract (offline generation rules), spec (offline: code
  covers every dependency / message / text / selection field in the design spec),
  fm (CALL FUNCTION params / sections / types vs the live FM over RFC — absorbed
  from the former sap-check-fm), syntax (compiler-level EDITOR_SYNTAX_CHECK via
  RFC, catching real errors offline before any GUI upload). Writes tab-delimited
  result files with fix advice; feed findings to /sap-fix-abap. SAP connection
  optional — offline skips type/sql/fm/syntax; live dimensions that can't run
  report COULD_NOT_CHECK, never a silent pass.
  Prerequisites: SAP GUI (SAP.Functions COM); SAP NCo 3.1 (32-bit) + the dev-init
  wrapper Z_GENERIC_RFC_WRAPPER_TBL for the syntax dimension.
argument-hint: "<path-to-abap-source-file> [--dimensions naming,type,sql,contract,spec,conv,fm,syntax] [--offline] [--gate]"

SAP Check ABAP Skill

You validate ABAP source code quality before deployment across several **dimensions**, then write findings that `/sap-fix-abap` can consume. This one skill supersedes the former `sap-check-abap` + `sap-check-fm` (now merged here, in sap-dev-core).

Task: $ARGUMENTS

---

Dimensions — what to run

The skill runs a set of **dimensions**. By default it runs every dimension that applies to the source and the connection state; narrow with `--dimensions a,b,…`, force offline with `--offline`, or use the fast pre-deploy preset `--gate` (= `syntax,fm`).

| Dimension | What it checks | Live? | Step | Output file | |---|---|---|---|---| | `naming` | variable + top-level object names | offline | 1.5 + 3 (Phase 5b naming) | `<file>.check.tsv` | | `type` | data-type validity via DDIC | RFC | 3 | `<file>.check.tsv` | | `sql` | SELECT/UPDATE/DELETE field validity | RFC | 3 | `<file>.check.tsv` | | `unused` | declared-but-unused variables | offline | 3 | `<file>.check.tsv` | | `contract` | generation-contract rules (§ codes) | offline | 3 | `<file>.check.tsv` | | `spec` | spec-coverage (deps/messages/texts/selection) | offline | 3.5 / 3.6 / 3.7 | `<file>.check.tsv` | | `conv` | CONVEXIT + CURR/QUAN reference | offline | 3.7 | `<file>.check.tsv` | | **`fm`** | **CALL FUNCTION signatures vs live FM defs** | **RFC** | **3.8** | `<file>.check_fm.tsv` | | **`syntax`** | **compiler-level ABAP syntax (all errors + warnings)** | **RFC + wrapper** | **3.9** | `<file>.syntax.tsv` |

**Rules:**

• The `naming`/`type`/`sql`/`unused`/`contract`/`spec`/`conv` dimensions all run in

the one Step 3 VBS + the Step 3.5–3.7 sidecars (they share `<file>.check.tsv`). The new `fm` and `syntax` dimensions run their own steps and write their own result files; **Step 4 reads all three** and presents them together.

• **`syntax` applicability:** self-contained `REPORT`/`PROGRAM` sources compile

standalone (`-Subc 1`). A **function-module `FUNCTION…ENDFUNCTION` fragment** or a **class / interface pool** is not standalone-compilable, so the dimension runs the engine's **`-Wrap` mode** (Strategy A, proven live S4D 2026-07-04): the fragment is re-presented as a self-contained program so its **body** is syntax-checked **pre-insert** with zero writes to SAP, and findings are line-mapped back to the original file. A signature too complex to model — or a bare **include** with no compilation unit — degrades to `SYNTAX_COULD_NOT_CHECK` (never a false-fail), deferring to the deploy skill's authoritative in-context Ctrl+F2 after an inactive insert (`/sap-se37`, `/sap-se24`). Every other dimension still runs.

• Any live dimension that cannot reach SAP reports `COULD_NOT_CHECK` (honest

tri-state — never rendered as a pass).

---

Step 0 — Resolve Work Directory

**Resolve `work_dir` via the env-aware helper** — do NOT take `work_dir` from a direct `settings.json` read (that ignores the `SAPDEV_AI_WORK_DIR` env var and `userconfig.json`). Use the `WORK_DIR=` value printed by:

powershell -NoProfile -ExecutionPolicy Bypass -Command ". '<SAP_DEV_CORE_SHARED_DIR>\scripts\sap_settings_lib.ps1'; . '<SAP_DEV_CORE_SHARED_DIR>\scripts\sap_connection_lib.ps1'; Write-Output ('WORK_DIR=' + (Get-SapWorkDir))"

The settings note below still applies to the OTHER keys.

**Settings reads/writes follow `<SAP_DEV_CORE_SHARED_DIR>/rules/settings_lookup.md`** — merge per-key on the `.value` field (env var → `settings.local.json` → `userconfig.json` → `settings.json`); non-per-connection writes go to `userconfig.json`. Resolve cross-plugin paths: 3 levels up from `<SKILL_DIR>`, then into `sap-dev-core\settings.json` and (if present) `sap-dev-core\settings.local.json`. Read `custom_url`.

| Setting | Default if blank | |---|---| | `work_dir` | `C:\sap_dev_work` | | `custom_url` | `{work_dir}\custom` |

Set `{WORK_TEMP}` = `{work_dir}\temp`

Ensure the temp directory exists:

cmd /c if not exist "{WORK_TEMP}" mkdir "{WORK_TEMP}"

Set `{RUN_TEMP}` = the per-run scratch dir (`Get-SapRunTemp` mints + creates `{work_dir}\temp\run_<id>`):

powershell -NoProfile -ExecutionPolicy Bypass -Command ". '<SAP_DEV_CORE_SHARED_DIR>\scripts\sap_connection_lib.ps1'; Write-Output ('RUN_TEMP=' + (Get-SapRunTemp))"

Per the CLAUDE.md "Two-bucket temp model" write this skill's generated scratch (`*_run.ps1` / `*_run.vbs` and the `_run.json` state) under `{RUN_TEMP}`; keep `{WORK_TEMP}` (base) only for `Get-SapCurrentSessionPath -WorkTemp`.

---

Step 0.5 — Start Logging

Start a structured log run. State file: `{RUN_TEMP}\sap_check_abap_run.json`. Best-effort.

powershell -ExecutionPolicy Bypass -File "<SAP_DEV_CORE_SHARED_DIR>\scripts\sap_log_helper.ps1" -Action start -StateFile "{RUN_TEMP}\sap_check_abap_run.json" -Skill sap-check-abap -ParamsJson "{\"aba