@plurnk/plurnk-execs-sqlite 0.1.5 → 0.1.6

package/docs/sqlite.md

@@ -0,0 +1,16 @@
+# sqlite
+
+Runs one SQL statement in-process via `node:sqlite` and writes the result to the `results` channel as `application/json` — ready for a jsonpath body-matcher.
+
+## Database target
+
+`EXEC[sqlite](./app.db):…` runs against the file `./app.db` (created if absent) — a persistent, host-mutating database. With **no target** it runs against an ephemeral `:memory:` database that's gone when the statement finishes (pure, auto-run). A directory is never a valid target and falls back to `:memory:`.
+
+## Query vs mutation
+
+The result shape is decided by the statement's columns, never by parsing the SQL:
+
+- **Row-returning** (SELECT, RETURNING, PRAGMA) → an array of row objects.
+- **Mutation** (INSERT / UPDATE / DELETE / CREATE) → `{ changes, lastInsertRowid }`.
+
+Large integers come back stringified (JSON can't hold a bigint). A SQL or open error closes `results` as `errored` with status 500 and a `sqlite_error` telemetry note.

package/package.json

@@ -1,6 +1,6 @@
 {
     "name": "@plurnk/plurnk-execs-sqlite",
-    "version": "0.1.5",
+    "version": "0.1.6",
     "description": "SQLite runtime executor for plurnk-service's exec scheme — runs the sqlite runtime tag in-process via node:sqlite, results as application/json.",
     "keywords": [
         "plurnk",
@@ -46,7 +46,8 @@
     },
     "files": [
         "dist/**/*",
-        "README.md"
+        "README.md",
+        "docs/**/*"
     ],
     "scripts": {
         "test:lint": "tsc --noEmit",