@superglue/cli 1.1.29 → 1.1.30

package/dist/skills/superglue/SKILL.md

@@ -272,7 +272,8 @@ sg mcp edit --id <serverId> --auth-mode creator_api_key
 
 ```bash
 sg run list --tool my-tool --status failed --limit 20
-sg run get <runId> --fetch-results
+sg run get <runId> --full
+sg run get <runId> --fetch-results  # alias for --full
 ```
 
 **File attachments:**
@@ -397,16 +398,14 @@ sourceData = {
 
 ### Naming Rules — Avoiding Key Collisions
 
-Everything in `sourceData` shares ONE flat namespace, and the runtime does not detect collisions — when two sources use the same key, one silently wins and the tool misbehaves with no error. Prevent collisions at design time:
+Everything in `sourceData` shares ONE flat namespace. Saving a tool rejects payload input keys and step ids that collide with it:
 
-Never use these as payload input keys or step ids:
+- `__files__`, `currentItem`, `sg_auth_email` — injected by the runtime
+- `page`, `offset`, `cursor`, `limit`, `pageSize` — rejected only when the tool has a paginated step, because they would shadow the live pagination counters; on non-paginated tools they are allowed but discouraged (a later edit that adds pagination will force a rename)
+- the tool's own step ids — the step result overwrites a same-named payload key
+- anything starting with `<systemId>_` for a system the tool uses — the entire prefix is reserved for namespaced credentials (`<systemId>_<credKey>`, `<systemId>_url`), even names that don't match an existing credential, because credentials added to the system later would silently shadow the input
 
-- `__files__`, `currentItem`, `sg_auth_email` — overwritten by the runtime
-- `page`, `offset`, `cursor`, `limit`, `pageSize` — a payload key with one of these names shadows the live pagination counters; paginated steps silently stop advancing past the first page
-- the tool's own step ids — the step result overwrites a same-named payload key for all subsequent steps
-- credential keys available to a step (`<systemId>_<credKey>`, `<systemId>_url`, and any runtime-supplied credential names) — the credential value silently replaces the payload value
-
-Safe naming: payload inputs get specific, domain-flavored names (`customerId`, `reportMonth`, `maxResults` — never `page`, `limit`, `cursor`, `data`); step ids are verb-prefixed camelCase (`fetchUsers`) and distinct from every payload key. If the user asks for an input named like a reserved key, use a safe alternative and explain why.
+Safe naming: payload inputs get specific, domain-flavored names (`customerId`, `reportMonth`, `maxResults` — prefer `maxResults` over `limit` even on non-paginated tools, and never prefix an input with a referenced system's id); step ids are verb-prefixed camelCase (`fetchUsers`) and distinct from every payload key. If the user asks for an input named like a reserved key, use a safe alternative and explain why.
 
 ### Step Result Envelopes
 
@@ -559,7 +558,7 @@ sg system create --name "My API" --url https://api.example.com \
 
 ### Draft Management
 
-Drafts live in `.superglue/drafts/<draftId>.json`. Created by `sg tool build`, updated by `sg tool edit`, deleted by `sg tool save`. Ephemeral — not persisted across sessions.
+Drafts live under the active CLI config directory: local `.superglue/drafts/` when a project config exists, otherwise `~/.superglue/drafts/`. Draft filenames are base64url-encoded draft IDs, so use the `draftPath` returned by `sg tool build` and `sg tool edit` when inspecting raw draft files. Created by `sg tool build`, updated by `sg tool edit`, deleted by `sg tool save`. Ephemeral — not persisted to the superglue API until saved.
 
 ### Deployment
 
@@ -602,7 +601,7 @@ curl -X POST "https://api.superglue.cloud/v1/hooks/{toolId}?token=$SUPERGLUE_API
 - **POST for read-only ops** — GraphQL queries via POST should have `modify: false`. Don't rely on HTTP method alone
 - **Saving without approval** — always present `sg tool run` results and ask before `sg tool save`
 - **`sg_auth_email` in scheduled runs** — not available in scheduler-triggered executions. Tools using `<<sg_auth_email>>` will fail
-- **Reserved key collisions** — payload inputs or step ids named after reserved/injected keys (see Naming Rules) fail silently with wrong values, never with an error
+- **Reserved key collisions** — payload inputs or step ids named after reserved/injected keys (see Naming Rules) are rejected at save time; pick a different name instead of retrying
 
 ---
 
@@ -612,7 +611,7 @@ For protocol-specific error recovery (HTTP, Postgres, MSSQL, Redis, GraphQL, SFT
 
 ### Tool build fails validation
 
-`sg tool build` returns a validation error. Read the message, fix the config, and re-run. Common causes: missing `id`, empty `steps`, missing `url` on request steps, missing `transformCode` on transform steps, `systemId` on the step object instead of `step.config.systemId`.
+`sg tool build` returns a validation error. Read the message, fix the config, and re-run. Common causes: missing `id`, empty `steps`, missing `url` on request steps, missing `transformCode` on transform steps, `systemId` on the step object instead of `step.config.systemId`, or a key name collision (see Naming Rules) — rename the flagged input or step id.
 
 ### Tool runs but a step fails
 
@@ -642,7 +641,6 @@ For protocol-specific error recovery (HTTP, Postgres, MSSQL, Redis, GraphQL, SFT
 2. Re-run with `--include-step-results` to see the exact envelope shape
 3. Paginated steps merge into a single `.data` field — do not `.map()` over them
 4. Verify `sourceData.currentItem` is only referenced inside looping steps (set by the step's own `dataSelector`, not upstream)
-5. If a step silently received the wrong value (no error thrown) or pagination stopped after one page, check for a key collision between payload inputs, step ids, credential keys, and reserved runtime keys — see Naming Rules. Fix by renaming the input, not by working around it
 
 ### Persistent failures
 

package/dist/skills/superglue/references/http.md

@@ -196,11 +196,14 @@ pagination: {
 | cursorBased | `<<cursor>>`                 | null      | extracted via cursorPath |
 | all         | `<<limit>>` / `<<pageSize>>` | pageSize  | —                        |
 
-**Stop conditions** — JS expression evaluated in sandbox. Receives `(response, pageInfo)`:
+**Stop conditions** — JS expression evaluated in sandbox. Receives `(response, pageInfo, sourceData)`:
 
 - `response.data` = parsed API response body
 - `response.headers` = response headers
-- `pageInfo = { page, offset, cursor, totalFetched }`
+- `pageInfo = { page, offset, cursor, totalFetched, mergedResult }`
+- `pageInfo.totalFetched` counts the merged paginated result including the current page
+- `pageInfo.mergedResult` is the accumulated paginated result including the current page
+- `sourceData` contains the current step input data, with payload fields at the root plus previous step results
 - Return `true` to STOP
 
 ```javascript
@@ -208,6 +211,7 @@ pagination: {
 "response.data.items.length === 0";
 "response.data.hasMore === false";
 "pageInfo.totalFetched >= 1000";
+"pageInfo.totalFetched >= sourceData.maxResults";
 ```
 
 **Cursor extraction:** `cursorPath` is a simple dot-separated property path (e.g. `meta.next_cursor`, `data.0.id`) resolved against the parsed response body. It is not a JSONPath expression — wildcards, filters, and `$.` prefixes are not supported. For property names containing `@` (e.g. OData `@odata.nextLink`), use the name directly as the path. Null/undefined cursor stops pagination.
@@ -233,7 +237,7 @@ When an HTTP step downloads a file (PDF, CSV, binary response), set `outputFile:
 
 ## Common Pitfalls
 
-- Hardcoding base URLs that match the system's base URL instead of using `<<systemId_url>>`
+- Hardcoding base URLs that match the system's base URL instead of using `<<systemId_url>>`. Check with `sg system call` what format the URL is and if it contains e.g. a trailing slash.
 - Forgetting auth headers entirely — nothing is injected automatically, including for OAuth systems (only token refresh is automatic)
 - Forgetting `outputFile: true` on steps that download files the user wants
 - Putting pagination variables in the config but not setting a `pagination` block, or vice versa

package/dist/skills/superglue/references/mssql.md

@@ -1,5 +1,7 @@
 # MSSQL
 
+**This skill is for Microsoft SQL Server and Azure SQL only.** For SAP ASE (Sybase), use the `odbc` skill — SAP ASE requires the ODBC strategy with `odbc://` URLs, not `mssql://`.
+
 ## Step Configuration
 
 ```typescript
@@ -44,17 +46,29 @@ Azure SQL usernames often include `@servername` suffix (e.g., `myuser@myserver`)
 mssql://myuser%40myserver:mypassword@myserver.database.windows.net:1433/mydatabase
 ```
 
+**Named instances** (common for on-prem servers, e.g. ProLaw's `SERVER\PROLAW`) — use the `instanceName` query parameter. Backslash syntax (`SERVER\PROLAW`) in the host is rejected with an error; always translate it to the query parameter:
+
+```
+mssql://user:password@SERVER/database?instanceName=PROLAW
+```
+
+When an instance name is set, the port is ignored — the instance's port is resolved via the SQL Browser service (UDP 1434), which must be running and reachable. If SQL Browser is unavailable (or the server is reached through the Secure Gateway tunnel, which forwards a single TCP port and cannot carry UDP), have the DBA assign a static TCP port to the instance and connect with `host:port` without an instance name instead.
+
 **Connection parameters** can be appended as query strings:
 
 ```
 mssql://user:password@host:port/database?encrypt=true&trustServerCertificate=false
 ```
 
-| Parameter                | Default | Notes                                   |
-| ------------------------ | ------- | --------------------------------------- |
-| `encrypt`                | `true`  | Required for Azure SQL                  |
-| `trustServerCertificate` | `false` | Whether to trust the server certificate |
-| `database`               | —       | Can be in URL path or query parameter   |
+| Parameter                | Default | Notes                                               |
+| ------------------------ | ------- | --------------------------------------------------- |
+| `encrypt`                | `true`  | Required for Azure SQL                              |
+| `trustServerCertificate` | `false` | Whether to trust the server certificate             |
+| `database`               | —       | Can be in URL path or query parameter               |
+| `instanceName`           | —       | SQL Server named instance (case-sensitive spelling) |
+| `domain`                 | —       | Windows domain — switches to NTLM authentication    |
+
+Unknown connection parameters cause an error rather than being silently ignored.
 
 ### Body Format
 
@@ -77,6 +91,24 @@ Credentials are embedded directly in the connection URL using `<<systemId_creden
 mssql://<<my_db_user>>:<<my_db_password>>@<<my_db_host>>:<<my_db_port>>/<<my_db_database>>
 ```
 
+### Windows Authentication (NTLM)
+
+For on-prem SQL Servers that only accept Windows domain logins (no SQL auth), store these credential keys on the system:
+
+| Credential Key | Required | Notes                                                 |
+| -------------- | -------- | ----------------------------------------------------- |
+| `authType`     | yes      | `ntlm`                                                |
+| `domain`       | yes      | Windows domain (e.g. `CORP`)                          |
+| `user`         | yes      | Windows username (also accepts `username`, or in URL) |
+| `password`     | yes      | Windows password (or in URL)                          |
+
+```
+URL:  mssql://<<sys_user>>:<<sys_password>>@<<sys_host>>:<<sys_port>>/<<sys_database>>
+Credentials: { authType: "ntlm", domain, user, password, host, port, database }
+```
+
+Alternatively, pass the domain as a `?domain=CORP` query parameter on the URL — any URL with a `domain` set authenticates via NTLM. Kerberos-only environments are not supported; ask the DBA to permit NTLM or create a SQL login.
+
 ### Azure Active Directory Authentication
 
 For Azure AD auth, store the following credential keys on the system alongside the standard connection fields:
@@ -91,7 +123,7 @@ For Azure AD auth, store the following credential keys on the system alongside t
 | `user`         | yes      | Azure AD username (also accepts `username`, or in URL) |
 | `password`     | yes      | Azure AD password (or in URL)                          |
 
-`user` and `password` can be stored as credentials or embedded in the URL — credentials take precedence.
+`user` and `password` can be stored as credentials or embedded in the URL — values resolved into the URL take precedence; stored credentials only fill gaps the URL omits.
 
 ```
 URL:  mssql://<<sys_user>>:<<sys_password>>@myserver.database.windows.net:1433/mydb
@@ -160,11 +192,14 @@ With OUTPUT (MSSQL equivalent of RETURNING):
 
 ### Return Value
 
-Returns the recordset — array of row objects with column names as keys.
+The shape depends on what the query produces:
 
-- `SELECT`: array of matching rows
+- `SELECT` (single recordset): array of row objects with column names as keys
 - `INSERT/UPDATE/DELETE` with `OUTPUT`: array of returned rows
-- Mutations without `OUTPUT`: empty array
+- Mutations without `OUTPUT`: `{ "rowsAffected": [n] }` — one count per statement
+- Stored procedures / batches returning multiple recordsets: array of recordset arrays (e.g. `[[...rows], [...rows]]`)
+
+Check the shape before indexing: a multi-recordset result is an array of arrays, not a flat row list.
 
 ### Batch Operations
 
@@ -182,6 +217,15 @@ Use a data selector returning an array to execute a query per item:
 - Idle pools cleaned up automatically
 - SSL/TLS encryption enabled by default for Azure SQL
 
+### On-Prem SQL Server (e.g. ProLaw, legacy ERP)
+
+Defaults are tuned for Azure SQL; on-prem servers usually need adjustments:
+
+- **Self-signed certificates** — on-prem servers rarely have CA-signed certs, so the default `trustServerCertificate=false` fails the TLS handshake. Add `?trustServerCertificate=true` (keeps encryption, skips CA validation).
+- **Old SQL Server versions** — SQL Server 2008/2008 R2 without TLS 1.2 patches cannot complete a modern TLS handshake. Use `?encrypt=false` as a last resort, preferably only over the Secure Gateway tunnel.
+- **Named instances** — use the `?instanceName=` query parameter (see URL Format above); remember SQL Browser (UDP 1434) is required for instance resolution and does not work through the Secure Gateway tunnel.
+- **Firewalled servers** — on-prem databases are typically not internet-reachable; configure the system with a Secure Gateway tunnel.
+
 ### Retry Behavior
 
 Default retries from server config. On final failure, error includes query text and params.
@@ -189,9 +233,12 @@ Default retries from server config. On final failure, error includes query text
 ## Common Pitfalls
 
 - Using `$1`, `$2` (PostgreSQL syntax) instead of `@param1`, `@param2` — MSSQL uses named `@param` parameters
-- Forgetting `OUTPUT INSERTED.*` on INSERT/UPDATE/DELETE when you need the result — without it the step returns an empty array
+- Forgetting `OUTPUT INSERTED.*` on INSERT/UPDATE/DELETE when you need the row data — without it the step returns `{ "rowsAffected": [n] }`, not rows
+- Treating a multi-recordset stored procedure result as a flat row array — it is an array of recordset arrays
 - Accessing step result directly instead of via `.data` — database results follow the same envelope as all steps: `sourceData.stepId.data`
 - Not URL-encoding `@` in Azure SQL usernames — `myuser@myserver` must be `myuser%40myserver` in the connection URL
+- Using a named instance through the Secure Gateway tunnel — instance resolution needs UDP 1434, which the tunnel does not forward; use a static TCP port instead
+- Forgetting `?trustServerCertificate=true` for on-prem servers with self-signed certificates
 
 ## Error Recovery
 
@@ -204,4 +251,8 @@ When an MSSQL step fails and the cause is not obvious:
    - Confirm the client IP is allowed in the Azure SQL firewall rules
    - Verify the username includes the `@servername` suffix if required, properly encoded as `%40`
    - Confirm `encrypt=true` is set (required for Azure SQL)
-5. **Verify parameter types** — MSSQL infers types from the values passed. Passing a string where a number is expected may cause implicit conversion errors. Ensure `params` values match expected column types.
+5. **Check on-prem connection issues** — if connecting to an on-prem server:
+   - TLS handshake errors (`self signed certificate`, `unable to verify`) → add `?trustServerCertificate=true`; for unpatched SQL Server 2008-era hosts, `?encrypt=false`
+   - `Failed to connect ... in 15000ms` with a named instance → SQL Browser (UDP 1434) is unreachable; use a static TCP port instead
+   - `Login failed` with Windows-auth-only servers → set `authType: "ntlm"` and `domain` in credentials, or ask the DBA for a SQL login
+6. **Verify parameter types** — MSSQL infers types from the values passed. Passing a string where a number is expected may cause implicit conversion errors. Ensure `params` values match expected column types.

package/dist/skills/superglue/references/superglue-info.md

@@ -51,18 +51,18 @@ Use this section as the source of truth for high-level web app navigation. Do no
 The persistent left sidebar contains these top-level items:
 
 - **Agent** (`/`) — primary AI chat assistant for building and debugging tools and systems.
-- **Landscape** (`/landscape`) — graph overview of tools, systems, and their connections.
 - **Tools** (`/tools`, detail at `/tools/{toolId}`) — saved tools; opens the tool playground for editing, testing, and running.
 - **Systems** (`/systems`, detail at `/systems/{systemId}`) — connected external systems with credentials and documentation.
 - **My Credentials** (`/credentials`) — credential management. The **My Credentials** section is for personal user-owned system credentials; the **superglue API keys** section is for API keys. Admins and members open it inside the normal app shell from the sidebar.
 - **Runs** (`/runs`, detail at `/runs/{runId}`) — execution history for full, draft, and single-step tool runs. Runs is a top-level sidebar item, not a Control Panel sub-item.
 - **Control Panel** (`/admin`) — expandable group for organization and account administration:
   - **Overview** (`/admin`) — dashboard summary, including failed-run and schedule summaries.
+  - **Landscape** (`/landscape`) — graph overview of tools, systems, and their connections.
   - **Schedules** (`/admin?view=schedules`) — scheduled and recurring tool runs.
   - **Access Rules** (`/admin/access`) — role and access-rule configuration; visible to admins on paid tiers. Use the access-rules reference for RBAC behavior, base roles, and personal roles.
-  - **Organization** (`/admin?view=organization`) — organization members, invites, and org-level settings; visible on paid tiers. Member management actions are admin-gated.
+  - **Organization** (`/admin?view=organization`) — visible on paid tiers, tabbed page: **Members** (members, invites, role assignment; management actions admin-gated), **Settings** (`?tab=settings`, admin-only: organization name and logo, run preferences like draft/single-step run visibility, run result storage, member run visibility, and delete-all-runs), and **Notifications** (`?tab=notifications`, admin-only: failure-alert channels for Slack and email).
   - **MCP Servers** (`/admin?view=mcp-servers`) — manage and export named MCP server endpoints. Use the MCP documentation for behavior, permissions, and client setup.
-  - **Notifications** (`/admin?view=notifications`) — user notification preferences and failure-alert channels.
+  - **Notifications** (`/admin?view=notifications`) — failure-alert channels; shown as a Control Panel sub-item only on the free tier. On paid tiers this lives in the Organization page's Notifications tab instead.
 
 Below the navigation, the sidebar shows the organization switcher and Sign Out. Depending on the plan, it also shows **Upgrade Plan** or **Subscription**; Subscription opens the external Stripe billing portal. There is no standalone Settings, Profile, or in-app Billing page.
 

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@superglue/cli",
-  "version": "1.1.29",
+  "version": "1.1.30",
   "bin": {
     "sg": "./dist/cli.js"
   },

package/skills/superglue/SKILL.md

@@ -272,7 +272,8 @@ sg mcp edit --id <serverId> --auth-mode creator_api_key
 
 ```bash
 sg run list --tool my-tool --status failed --limit 20
-sg run get <runId> --fetch-results
+sg run get <runId> --full
+sg run get <runId> --fetch-results  # alias for --full
 ```
 
 **File attachments:**
@@ -397,16 +398,14 @@ sourceData = {
 
 ### Naming Rules — Avoiding Key Collisions
 
-Everything in `sourceData` shares ONE flat namespace, and the runtime does not detect collisions — when two sources use the same key, one silently wins and the tool misbehaves with no error. Prevent collisions at design time:
+Everything in `sourceData` shares ONE flat namespace. Saving a tool rejects payload input keys and step ids that collide with it:
 
-Never use these as payload input keys or step ids:
+- `__files__`, `currentItem`, `sg_auth_email` — injected by the runtime
+- `page`, `offset`, `cursor`, `limit`, `pageSize` — rejected only when the tool has a paginated step, because they would shadow the live pagination counters; on non-paginated tools they are allowed but discouraged (a later edit that adds pagination will force a rename)
+- the tool's own step ids — the step result overwrites a same-named payload key
+- anything starting with `<systemId>_` for a system the tool uses — the entire prefix is reserved for namespaced credentials (`<systemId>_<credKey>`, `<systemId>_url`), even names that don't match an existing credential, because credentials added to the system later would silently shadow the input
 
-- `__files__`, `currentItem`, `sg_auth_email` — overwritten by the runtime
-- `page`, `offset`, `cursor`, `limit`, `pageSize` — a payload key with one of these names shadows the live pagination counters; paginated steps silently stop advancing past the first page
-- the tool's own step ids — the step result overwrites a same-named payload key for all subsequent steps
-- credential keys available to a step (`<systemId>_<credKey>`, `<systemId>_url`, and any runtime-supplied credential names) — the credential value silently replaces the payload value
-
-Safe naming: payload inputs get specific, domain-flavored names (`customerId`, `reportMonth`, `maxResults` — never `page`, `limit`, `cursor`, `data`); step ids are verb-prefixed camelCase (`fetchUsers`) and distinct from every payload key. If the user asks for an input named like a reserved key, use a safe alternative and explain why.
+Safe naming: payload inputs get specific, domain-flavored names (`customerId`, `reportMonth`, `maxResults` — prefer `maxResults` over `limit` even on non-paginated tools, and never prefix an input with a referenced system's id); step ids are verb-prefixed camelCase (`fetchUsers`) and distinct from every payload key. If the user asks for an input named like a reserved key, use a safe alternative and explain why.
 
 ### Step Result Envelopes
 
@@ -559,7 +558,7 @@ sg system create --name "My API" --url https://api.example.com \
 
 ### Draft Management
 
-Drafts live in `.superglue/drafts/<draftId>.json`. Created by `sg tool build`, updated by `sg tool edit`, deleted by `sg tool save`. Ephemeral — not persisted across sessions.
+Drafts live under the active CLI config directory: local `.superglue/drafts/` when a project config exists, otherwise `~/.superglue/drafts/`. Draft filenames are base64url-encoded draft IDs, so use the `draftPath` returned by `sg tool build` and `sg tool edit` when inspecting raw draft files. Created by `sg tool build`, updated by `sg tool edit`, deleted by `sg tool save`. Ephemeral — not persisted to the superglue API until saved.
 
 ### Deployment
 
@@ -602,7 +601,7 @@ curl -X POST "https://api.superglue.cloud/v1/hooks/{toolId}?token=$SUPERGLUE_API
 - **POST for read-only ops** — GraphQL queries via POST should have `modify: false`. Don't rely on HTTP method alone
 - **Saving without approval** — always present `sg tool run` results and ask before `sg tool save`
 - **`sg_auth_email` in scheduled runs** — not available in scheduler-triggered executions. Tools using `<<sg_auth_email>>` will fail
-- **Reserved key collisions** — payload inputs or step ids named after reserved/injected keys (see Naming Rules) fail silently with wrong values, never with an error
+- **Reserved key collisions** — payload inputs or step ids named after reserved/injected keys (see Naming Rules) are rejected at save time; pick a different name instead of retrying
 
 ---
 
@@ -612,7 +611,7 @@ For protocol-specific error recovery (HTTP, Postgres, MSSQL, Redis, GraphQL, SFT
 
 ### Tool build fails validation
 
-`sg tool build` returns a validation error. Read the message, fix the config, and re-run. Common causes: missing `id`, empty `steps`, missing `url` on request steps, missing `transformCode` on transform steps, `systemId` on the step object instead of `step.config.systemId`.
+`sg tool build` returns a validation error. Read the message, fix the config, and re-run. Common causes: missing `id`, empty `steps`, missing `url` on request steps, missing `transformCode` on transform steps, `systemId` on the step object instead of `step.config.systemId`, or a key name collision (see Naming Rules) — rename the flagged input or step id.
 
 ### Tool runs but a step fails
 
@@ -642,7 +641,6 @@ For protocol-specific error recovery (HTTP, Postgres, MSSQL, Redis, GraphQL, SFT
 2. Re-run with `--include-step-results` to see the exact envelope shape
 3. Paginated steps merge into a single `.data` field — do not `.map()` over them
 4. Verify `sourceData.currentItem` is only referenced inside looping steps (set by the step's own `dataSelector`, not upstream)
-5. If a step silently received the wrong value (no error thrown) or pagination stopped after one page, check for a key collision between payload inputs, step ids, credential keys, and reserved runtime keys — see Naming Rules. Fix by renaming the input, not by working around it
 
 ### Persistent failures
 

package/skills/superglue/references/http.md

@@ -196,11 +196,14 @@ pagination: {
 | cursorBased | `<<cursor>>`                 | null      | extracted via cursorPath |
 | all         | `<<limit>>` / `<<pageSize>>` | pageSize  | —                        |
 
-**Stop conditions** — JS expression evaluated in sandbox. Receives `(response, pageInfo)`:
+**Stop conditions** — JS expression evaluated in sandbox. Receives `(response, pageInfo, sourceData)`:
 
 - `response.data` = parsed API response body
 - `response.headers` = response headers
-- `pageInfo = { page, offset, cursor, totalFetched }`
+- `pageInfo = { page, offset, cursor, totalFetched, mergedResult }`
+- `pageInfo.totalFetched` counts the merged paginated result including the current page
+- `pageInfo.mergedResult` is the accumulated paginated result including the current page
+- `sourceData` contains the current step input data, with payload fields at the root plus previous step results
 - Return `true` to STOP
 
 ```javascript
@@ -208,6 +211,7 @@ pagination: {
 "response.data.items.length === 0";
 "response.data.hasMore === false";
 "pageInfo.totalFetched >= 1000";
+"pageInfo.totalFetched >= sourceData.maxResults";
 ```
 
 **Cursor extraction:** `cursorPath` is a simple dot-separated property path (e.g. `meta.next_cursor`, `data.0.id`) resolved against the parsed response body. It is not a JSONPath expression — wildcards, filters, and `$.` prefixes are not supported. For property names containing `@` (e.g. OData `@odata.nextLink`), use the name directly as the path. Null/undefined cursor stops pagination.
@@ -233,7 +237,7 @@ When an HTTP step downloads a file (PDF, CSV, binary response), set `outputFile:
 
 ## Common Pitfalls
 
-- Hardcoding base URLs that match the system's base URL instead of using `<<systemId_url>>`
+- Hardcoding base URLs that match the system's base URL instead of using `<<systemId_url>>`. Check with `sg system call` what format the URL is and if it contains e.g. a trailing slash.
 - Forgetting auth headers entirely — nothing is injected automatically, including for OAuth systems (only token refresh is automatic)
 - Forgetting `outputFile: true` on steps that download files the user wants
 - Putting pagination variables in the config but not setting a `pagination` block, or vice versa

package/skills/superglue/references/mssql.md

@@ -1,5 +1,7 @@
 # MSSQL
 
+**This skill is for Microsoft SQL Server and Azure SQL only.** For SAP ASE (Sybase), use the `odbc` skill — SAP ASE requires the ODBC strategy with `odbc://` URLs, not `mssql://`.
+
 ## Step Configuration
 
 ```typescript
@@ -44,17 +46,29 @@ Azure SQL usernames often include `@servername` suffix (e.g., `myuser@myserver`)
 mssql://myuser%40myserver:mypassword@myserver.database.windows.net:1433/mydatabase
 ```
 
+**Named instances** (common for on-prem servers, e.g. ProLaw's `SERVER\PROLAW`) — use the `instanceName` query parameter. Backslash syntax (`SERVER\PROLAW`) in the host is rejected with an error; always translate it to the query parameter:
+
+```
+mssql://user:password@SERVER/database?instanceName=PROLAW
+```
+
+When an instance name is set, the port is ignored — the instance's port is resolved via the SQL Browser service (UDP 1434), which must be running and reachable. If SQL Browser is unavailable (or the server is reached through the Secure Gateway tunnel, which forwards a single TCP port and cannot carry UDP), have the DBA assign a static TCP port to the instance and connect with `host:port` without an instance name instead.
+
 **Connection parameters** can be appended as query strings:
 
 ```
 mssql://user:password@host:port/database?encrypt=true&trustServerCertificate=false
 ```
 
-| Parameter                | Default | Notes                                   |
-| ------------------------ | ------- | --------------------------------------- |
-| `encrypt`                | `true`  | Required for Azure SQL                  |
-| `trustServerCertificate` | `false` | Whether to trust the server certificate |
-| `database`               | —       | Can be in URL path or query parameter   |
+| Parameter                | Default | Notes                                               |
+| ------------------------ | ------- | --------------------------------------------------- |
+| `encrypt`                | `true`  | Required for Azure SQL                              |
+| `trustServerCertificate` | `false` | Whether to trust the server certificate             |
+| `database`               | —       | Can be in URL path or query parameter               |
+| `instanceName`           | —       | SQL Server named instance (case-sensitive spelling) |
+| `domain`                 | —       | Windows domain — switches to NTLM authentication    |
+
+Unknown connection parameters cause an error rather than being silently ignored.
 
 ### Body Format
 
@@ -77,6 +91,24 @@ Credentials are embedded directly in the connection URL using `<<systemId_creden
 mssql://<<my_db_user>>:<<my_db_password>>@<<my_db_host>>:<<my_db_port>>/<<my_db_database>>
 ```
 
+### Windows Authentication (NTLM)
+
+For on-prem SQL Servers that only accept Windows domain logins (no SQL auth), store these credential keys on the system:
+
+| Credential Key | Required | Notes                                                 |
+| -------------- | -------- | ----------------------------------------------------- |
+| `authType`     | yes      | `ntlm`                                                |
+| `domain`       | yes      | Windows domain (e.g. `CORP`)                          |
+| `user`         | yes      | Windows username (also accepts `username`, or in URL) |
+| `password`     | yes      | Windows password (or in URL)                          |
+
+```
+URL:  mssql://<<sys_user>>:<<sys_password>>@<<sys_host>>:<<sys_port>>/<<sys_database>>
+Credentials: { authType: "ntlm", domain, user, password, host, port, database }
+```
+
+Alternatively, pass the domain as a `?domain=CORP` query parameter on the URL — any URL with a `domain` set authenticates via NTLM. Kerberos-only environments are not supported; ask the DBA to permit NTLM or create a SQL login.
+
 ### Azure Active Directory Authentication
 
 For Azure AD auth, store the following credential keys on the system alongside the standard connection fields:
@@ -91,7 +123,7 @@ For Azure AD auth, store the following credential keys on the system alongside t
 | `user`         | yes      | Azure AD username (also accepts `username`, or in URL) |
 | `password`     | yes      | Azure AD password (or in URL)                          |
 
-`user` and `password` can be stored as credentials or embedded in the URL — credentials take precedence.
+`user` and `password` can be stored as credentials or embedded in the URL — values resolved into the URL take precedence; stored credentials only fill gaps the URL omits.
 
 ```
 URL:  mssql://<<sys_user>>:<<sys_password>>@myserver.database.windows.net:1433/mydb
@@ -160,11 +192,14 @@ With OUTPUT (MSSQL equivalent of RETURNING):
 
 ### Return Value
 
-Returns the recordset — array of row objects with column names as keys.
+The shape depends on what the query produces:
 
-- `SELECT`: array of matching rows
+- `SELECT` (single recordset): array of row objects with column names as keys
 - `INSERT/UPDATE/DELETE` with `OUTPUT`: array of returned rows
-- Mutations without `OUTPUT`: empty array
+- Mutations without `OUTPUT`: `{ "rowsAffected": [n] }` — one count per statement
+- Stored procedures / batches returning multiple recordsets: array of recordset arrays (e.g. `[[...rows], [...rows]]`)
+
+Check the shape before indexing: a multi-recordset result is an array of arrays, not a flat row list.
 
 ### Batch Operations
 
@@ -182,6 +217,15 @@ Use a data selector returning an array to execute a query per item:
 - Idle pools cleaned up automatically
 - SSL/TLS encryption enabled by default for Azure SQL
 
+### On-Prem SQL Server (e.g. ProLaw, legacy ERP)
+
+Defaults are tuned for Azure SQL; on-prem servers usually need adjustments:
+
+- **Self-signed certificates** — on-prem servers rarely have CA-signed certs, so the default `trustServerCertificate=false` fails the TLS handshake. Add `?trustServerCertificate=true` (keeps encryption, skips CA validation).
+- **Old SQL Server versions** — SQL Server 2008/2008 R2 without TLS 1.2 patches cannot complete a modern TLS handshake. Use `?encrypt=false` as a last resort, preferably only over the Secure Gateway tunnel.
+- **Named instances** — use the `?instanceName=` query parameter (see URL Format above); remember SQL Browser (UDP 1434) is required for instance resolution and does not work through the Secure Gateway tunnel.
+- **Firewalled servers** — on-prem databases are typically not internet-reachable; configure the system with a Secure Gateway tunnel.
+
 ### Retry Behavior
 
 Default retries from server config. On final failure, error includes query text and params.
@@ -189,9 +233,12 @@ Default retries from server config. On final failure, error includes query text
 ## Common Pitfalls
 
 - Using `$1`, `$2` (PostgreSQL syntax) instead of `@param1`, `@param2` — MSSQL uses named `@param` parameters
-- Forgetting `OUTPUT INSERTED.*` on INSERT/UPDATE/DELETE when you need the result — without it the step returns an empty array
+- Forgetting `OUTPUT INSERTED.*` on INSERT/UPDATE/DELETE when you need the row data — without it the step returns `{ "rowsAffected": [n] }`, not rows
+- Treating a multi-recordset stored procedure result as a flat row array — it is an array of recordset arrays
 - Accessing step result directly instead of via `.data` — database results follow the same envelope as all steps: `sourceData.stepId.data`
 - Not URL-encoding `@` in Azure SQL usernames — `myuser@myserver` must be `myuser%40myserver` in the connection URL
+- Using a named instance through the Secure Gateway tunnel — instance resolution needs UDP 1434, which the tunnel does not forward; use a static TCP port instead
+- Forgetting `?trustServerCertificate=true` for on-prem servers with self-signed certificates
 
 ## Error Recovery
 
@@ -204,4 +251,8 @@ When an MSSQL step fails and the cause is not obvious:
    - Confirm the client IP is allowed in the Azure SQL firewall rules
    - Verify the username includes the `@servername` suffix if required, properly encoded as `%40`
    - Confirm `encrypt=true` is set (required for Azure SQL)
-5. **Verify parameter types** — MSSQL infers types from the values passed. Passing a string where a number is expected may cause implicit conversion errors. Ensure `params` values match expected column types.
+5. **Check on-prem connection issues** — if connecting to an on-prem server:
+   - TLS handshake errors (`self signed certificate`, `unable to verify`) → add `?trustServerCertificate=true`; for unpatched SQL Server 2008-era hosts, `?encrypt=false`
+   - `Failed to connect ... in 15000ms` with a named instance → SQL Browser (UDP 1434) is unreachable; use a static TCP port instead
+   - `Login failed` with Windows-auth-only servers → set `authType: "ntlm"` and `domain` in credentials, or ask the DBA for a SQL login
+6. **Verify parameter types** — MSSQL infers types from the values passed. Passing a string where a number is expected may cause implicit conversion errors. Ensure `params` values match expected column types.

package/skills/superglue/references/superglue-info.md

@@ -51,18 +51,18 @@ Use this section as the source of truth for high-level web app navigation. Do no
 The persistent left sidebar contains these top-level items:
 
 - **Agent** (`/`) — primary AI chat assistant for building and debugging tools and systems.
-- **Landscape** (`/landscape`) — graph overview of tools, systems, and their connections.
 - **Tools** (`/tools`, detail at `/tools/{toolId}`) — saved tools; opens the tool playground for editing, testing, and running.
 - **Systems** (`/systems`, detail at `/systems/{systemId}`) — connected external systems with credentials and documentation.
 - **My Credentials** (`/credentials`) — credential management. The **My Credentials** section is for personal user-owned system credentials; the **superglue API keys** section is for API keys. Admins and members open it inside the normal app shell from the sidebar.
 - **Runs** (`/runs`, detail at `/runs/{runId}`) — execution history for full, draft, and single-step tool runs. Runs is a top-level sidebar item, not a Control Panel sub-item.
 - **Control Panel** (`/admin`) — expandable group for organization and account administration:
   - **Overview** (`/admin`) — dashboard summary, including failed-run and schedule summaries.
+  - **Landscape** (`/landscape`) — graph overview of tools, systems, and their connections.
   - **Schedules** (`/admin?view=schedules`) — scheduled and recurring tool runs.
   - **Access Rules** (`/admin/access`) — role and access-rule configuration; visible to admins on paid tiers. Use the access-rules reference for RBAC behavior, base roles, and personal roles.
-  - **Organization** (`/admin?view=organization`) — organization members, invites, and org-level settings; visible on paid tiers. Member management actions are admin-gated.
+  - **Organization** (`/admin?view=organization`) — visible on paid tiers, tabbed page: **Members** (members, invites, role assignment; management actions admin-gated), **Settings** (`?tab=settings`, admin-only: organization name and logo, run preferences like draft/single-step run visibility, run result storage, member run visibility, and delete-all-runs), and **Notifications** (`?tab=notifications`, admin-only: failure-alert channels for Slack and email).
   - **MCP Servers** (`/admin?view=mcp-servers`) — manage and export named MCP server endpoints. Use the MCP documentation for behavior, permissions, and client setup.
-  - **Notifications** (`/admin?view=notifications`) — user notification preferences and failure-alert channels.
+  - **Notifications** (`/admin?view=notifications`) — failure-alert channels; shown as a Control Panel sub-item only on the free tier. On paid tiers this lives in the Organization page's Notifications tab instead.
 
 Below the navigation, the sidebar shows the organization switcher and Sign Out. Depending on the plan, it also shows **Upgrade Plan** or **Subscription**; Subscription opens the external Stripe billing portal. There is no standalone Settings, Profile, or in-app Billing page.