@celigo/api-specs 0.2.131 → 0.2.133

package/dist/ai-agent.yml

@@ -158,7 +158,7 @@ paths:
                   aiAgent:
                     provider: gemini
                     litellm:
-                      model: gemini/gemini-2.5-flash
+                      model: gemini-2.5-flash
                       maxCompletionTokens: 2000
                       _overrides:
                         gemini:
@@ -839,7 +839,8 @@ components:
     AiAgentConfig:
       type: object
       description: |
-        AI Agent configuration used by both AiAgentImport and GuardrailImport (ai_agent type).
+        AI agent configuration for an AiAgentImport. An `ai_agent` guardrail is configured by
+        the guardrail spec's `GuardrailAiAgentConfig` instead, which has no `tools` arrays.
 
         Configures which AI provider and model to use, along with instructions, parameter
         tuning, output format, and available tools. Providers come in two families with
@@ -1112,7 +1113,6 @@ components:
                 - gpt-5-mini
                 - gpt-5-nano
                 - gpt-4.1-mini
-                - gpt-4o
               x-celigo-ui-override: Required by the AI agent configuration form. Encoded here to mirror the form so builders produce connectable configurations.
               x-celigo-ai-guidance:
                 - |-
@@ -1192,7 +1192,9 @@ components:
               minimum: 100
               maximum: 128000
               default: 5000
-              description: Maximum number of tokens in the model's response (server default observed live on create)
+              description: |-
+                Caps the length of the model's response, in tokens. An agent created without it is
+                saved with 5000; raise it for long outputs such as large extractions.
             serviceTier:
               type: string
               enum:
@@ -5787,7 +5789,8 @@ components:
             - filter
           description: |-
             Filter applied to incoming records before they are sent to the LLM. Records that match
-            continue through the agent; records that don't are silently dropped.
+            continue through the agent; records that don't skip it, counted as ignored with no model
+            call, and continue to the next step.
           allOf:
             - $ref: '#/components/schemas/Filter'
         responseTransform:

package/dist/export.yml

@@ -6196,7 +6196,7 @@ components:
               (`http.paging.body` likewise from its paging fields). A body that is not parseable JSON at
               rest — Handlebars block helpers outside a string value — renders every field EMPTY in that
               form, so keep templated variables inside a JSON string value:
-              `"variables": "{\"after\": {{#if export.http.paging.token}}\"{{export.http.paging.token}}\"{{else}}null{{/if}}}"`.
+              `"variables": "{\"after\": {{#if export.http.paging.token}}\"{{export.http.paging.token}}\"{{else}}null{{/if}} }"`.
             - |-
               The GraphQL form has no relative URI field: the connection `baseURI` is expected to be the
               GraphQL endpoint (a saved `relativeURI` still applies but is not shown). Create GraphQL

package/dist/import.yml

@@ -549,6 +549,10 @@ paths:
         composed, not delivered — an FTP import against an unreachable host
         still returns `{name, fileLocation}`.
 
+        The import runs with its own `settings` only: hooks receive
+        `options.settings` with an `import` key but no `integration` or `flow`
+        key, and no `_flowId` or `_integrationId`.
+
         The import must be saved; to run an import definition that is not saved
         (or a modified copy of one), use the virtual import
         `POST /v1/connections/{_id}/import`, which takes the import document in
@@ -1303,7 +1307,7 @@ components:
               ## EXAMPLES:
               - `{ "UserId": "12345", "Name": "John Doe", "Age": 30, "Preferences": { "Language": "en", "Notifications": true } }`
           examples:
-            - '{"id": "{{record.id}}", "name": "{{record.name}}", "total": {{record.total}}}'
+            - '{"id": "{{record.id}}", "name": "{{record.name}}", "total": {{record.total}} }'
         updateExpression:
           type: string
           x-celigo-ui-override: Required by the DynamoDB import form when method is updateItem (dynamodb.updateExpression required, visibleWhen method=updateItem).
@@ -1966,7 +1970,10 @@ components:
               the Import's mapping step runs first and shapes the source record,
               then this template renders against that mapped record.
               Reference the mapped record with the ``record`` root when wrapping it (e.g.
-              ``{"requestData": {{{jsonSerialize record}}}}``).
+              ``{"requestData": {{{jsonSerialize record}}} }``).
+              Keep a literal ``}`` apart from a closing ``}}`` or ``}}}`` (a space is enough): the save
+              accepts ``{{{jsonSerialize record}}}}``, but every record then fails with
+              ``handlebars_template_parse_error``.
               Avoid bare ``{{{jsonSerialize record}}}`` with no wrapper — leave the field undefined instead.
               Also avoid wrapping a single post-mapped scalar field in ``jsonSerialize`` (e.g.
               ``{{{jsonSerialize record.<scalar>}}}``) — that pattern is brittle.
@@ -2029,9 +2036,9 @@ components:
 
               Typical shapes:
               - Message-bus envelope:
-                ``{"MessageBody": {"class": "EntityUpdate", "payload": {{{jsonSerialize record}}}}}``
+                ``{"MessageBody": {"class": "EntityUpdate", "payload": {{{jsonSerialize record}}} }}``
               - Metadata wrap:
-                ``{"batchNumber": "{{timestamp dateTimeFormat CST}}", "interfaceType": "...", "requestData": {{{jsonSerialize record}}}}``
+                ``{"batchNumber": "{{timestamp dateTimeFormat CST}}", "interfaceType": "...", "requestData": {{{jsonSerialize record}}} }``
               - Named single-record array wrap (when destination wants ``[record]`` but each request sends one):
                 ``{"products": [{{{jsonSerialize record}}}]}``
 
@@ -2072,7 +2079,7 @@ components:
               current ``{{#each data}}`` item.
 
               Example shape:
-              ``{"header": {"sellerId": "{{connection.http.unencrypted.sellerId}}"}, "messages": [{{#each data}}{"sku": "{{sku}}", "qty": {{quantity}}}{{/each}}]}``
+              ``{"header": {"sellerId": "{{connection.http.unencrypted.sellerId}}"}, "messages": [{{#each data}}{"sku": "{{sku}}", "qty": {{quantity}} }{{/each}}]}``
             - |-
               ## Anti-pattern — DO NOT wrap a post-mapped scalar in jsonSerialize
 
@@ -2157,7 +2164,7 @@ components:
                 mix the two shapes in one body.
           examples:
             - - '{"name": "{{record.name}}", "email": "{{record.email}}"}'
-            - - '{"requestData": {{{jsonSerialize record}}}}'
+            - - '{"requestData": {{{jsonSerialize record}}} }'
             - - '{"values": [{{#each batch_of_records}}{{{jsonSerialize this.record}}}{{#unless @last}},{{/unless}}{{/each}}]}'
           x-celigo-agent:
             - handlebars
@@ -2647,9 +2654,7 @@ components:
 
             The referenced helper bundles the polling config plus a required status
             export (polled to check progress) and a result export (fetches the
-            final payload). An import configured with an async helper cannot carry
-            its own transform, output filter, or preSavePage hook — put that
-            processing in the result export instead.
+            final payload).
           examples:
             - 60a2c4e6f321d800129a1a3c
           x-celigo-agent:
@@ -2733,7 +2738,7 @@ components:
 
               - **Static** — only ``map`` + optional ``default``.  No HTTP
                 request issued.  Use for small controlled vocabularies.
-              - **Dynamic** — ``method`` + ``relativeURI`` (+ ``postBody``,
+              - **Dynamic** — ``method`` + ``relativeURI`` (+ ``body``,
                 ``headers``, ``extract``).  Issues a live HTTP request per
                 lookup.  Adds one HTTP request per record at runtime — factor
                 this into flow design for high-volume syncs.
@@ -2771,14 +2776,18 @@ components:
                   inject values from the incoming record — e.g.
                   ``/customers?email={{{email}}}`` or
                   ``/accounts/{{{accountId}}}``.
-              postBody:
+              body:
                 type: string
                 x-celigo-agent:
                   - handlebars
                 description: |
-                  Request body template for POST / PUT / PATCH lookup methods.
-                  Typically a JSON string with Handlebars placeholders —
-                  ``{"email":"{{{email}}}"}``. Ignored for GET / DELETE.
+                  Request body template for POST / PUT / PATCH lookup methods, typically
+                  a JSON string with Handlebars placeholders (``{"email":"{{{email}}}"}``);
+                  ignored for GET / DELETE. The record's fields sit at the top level:
+                  ``{{email}}`` renders the record's value, while ``{{record.email}}``
+                  renders empty.
+                x-celigo-ai-guidance:
+                  - 'Put the lookup''s request template here, not in ``postBody``: a ``postBody`` saves without an error but is not kept under ``http.lookups``, so the lookup goes out with no body and the run reports no error.'
               headers:
                 type: array
                 items:
@@ -4429,9 +4438,9 @@ components:
                                 line) to the rows that populate the
                                 subrecord.  Typically `$` when the
                                 subrecord data is carried on the same
-                                line item (the enclosing
-                                `lists[].jsonPath` has already selected
-                                the line).  Defaults to `$`.
+                                line item (iteration over the sublist
+                                has already selected the line).
+                                Defaults to `$`.
                             mapping:
                               type: object
                               description: |
@@ -5155,9 +5164,10 @@ components:
                   SQL `SELECT` statement that resolves the lookup value.
                   Typically returns a single row with a single column; use
                   `extract` to pick the column when the query returns multiple.
-                  Inject values from the incoming record with Handlebars — RDBMS
-                  queries use triple braces and the `record.` prefix:
-                  `SELECT id FROM users WHERE email = '{{{record.email}}}'`.
+                  Inject values from the incoming record with Handlebars and the
+                  `record.` prefix — double braces for strings, which the platform
+                  quotes and escapes, and triple braces for numbers:
+                  `SELECT id FROM users WHERE email = {{record.email}}`.
               extract:
                 type: string
                 description: |
@@ -5205,7 +5215,7 @@ components:
           description: |-
             SQL statements to execute; required when queryType is `["per_record"]`, `["per_page"]`, or `["first_page"]`, and typically contains a single statement.
             Each element is a complete SQL statement as a plain string — never a single bare string and never an array of objects (objects cause a Cast error).
-            Inject values from the incoming record with Handlebars using the `record.` prefix and triple braces, e.g. `'{{{record.name}}}'` for strings (single-quoted in the SQL) and `{{{record.quantity}}}` for numbers.
+            Inject values from the incoming record with Handlebars using the `record.` prefix: double braces for strings and dates (`{{record.name}}`; the platform quotes and escapes the value) and triple braces for numbers (`{{{record.quantity}}}`).
             Rendering is strict: a reference the render model cannot resolve (a bare `{{name}}` without prefix, or `record.*` when the arriving record is a grouped array) fails the step with `cannot_evaluate_handlebars` — it does not render an empty value.
             `record` and `rows` are aliases chosen by the arriving data's shape: an object record binds `record` (`rows` undefined); a grouped array-record (e.g. upstream `groupByFields`) binds `rows` (`record` undefined) — address grouped rows as `{{rows.0.field}}` or iterate `{{#each rows}}`.
             Double braces (`{{record.field}}`) output the value wrapped in single quotes with embedded quotes doubled, while triple braces output it raw; block helpers such as `{{#each}}` and `{{#if}}` use double braces as normal.
@@ -5219,7 +5229,7 @@ components:
               - CORRECT: ["INSERT INTO users (name, email) VALUES ({{record.name}}, {{record.email}})"]
               - WRONG: "INSERT INTO users ..."  (not an array)
               - WRONG: [{"query": "INSERT INTO users ..."}]  (objects are invalid — causes Cast error)
-              - WRONG: ["INSERT INTO users (name) VALUES ('{{name}}')"]  (missing record. prefix — rendering is strict, so the step fails with cannot_evaluate_handlebars)
+              - WRONG: ["INSERT INTO users (name) VALUES ({{name}})"]  (missing record. prefix — rendering is strict, so the step fails with cannot_evaluate_handlebars)
             - |-
               ## RDBMS HANDLEBARS SYNTAX
 
@@ -5253,7 +5263,7 @@ components:
                 | Numeric column value (INT, DECIMAL, ...) | `{{{record.field}}}`    |
                 | Stored procedure raw arg (already typed) | `{{{record.field}}}`    |
                 | Identifier (table/column name)           | `{{{record.field}}}`    |
-                | JSON literal (`PARSE_JSON`/JSONB)      | `{{{jsonSerialize record.subobject}}}` |
+                | JSON literal (`PARSE_JSON`/JSONB)      | `{{jsonSerialize record.subobject}}` (renders the JSON as a quoted string) |
 
               - **Block helpers** (`{{#each}}`, `{{#if}}`, `{{/each}}`,
                 `{{#unless}}`) use double braces as normal — they emit no value.
@@ -5272,7 +5282,7 @@ components:
               - Per-page batch (different Handlebars context — see `queryType: per_page` below): `["{{#each batch_of_records}}UPDATE staging SET done=1 WHERE id={{{this.record.id}}};{{/each}}"]`
           examples:
             - - INSERT INTO orders (id, total) VALUES ({{{record.id}}}, {{{record.total}}})
-            - - '{{#each batch_of_records}}UPDATE customers SET name=''{{{record.name}}}'' WHERE id={{{record.id}}};{{/each}}'
+            - - '{{#each batch_of_records}}UPDATE customers SET name={{record.name}} WHERE id={{{record.id}}};{{/each}}'
         queryType:
           type:
             - array
@@ -5301,7 +5311,7 @@ components:
             Prefer the highest-performance type the operation supports: bulk_load (Snowflake, Google BigQuery, Microsoft SQL Server on a cloud connection, Azure Synapse and Amazon Redshift — handles INSERT, upsert via `bulkLoad.primaryKeys`, and custom merge or ignore-existing logic via `bulkLoad.overrideMergeQuery`), then bulk_insert for pure INSERTs, then per_page, then per_record only when the logic cannot be expressed as a batch operation.
             For UPDATE, UPSERT, or ignore-existing logic on databases without bulk_load support, use per_page or per_record with `query` — bulk_insert has no duplicate-checking logic.
             Set a single value; do not combine types or use the legacy INSERT and UPDATE values.
-            per_page statements render against the whole batch (`batch_of_records`) rather than a single record — loop with `{{#each batch_of_records}}...{{{record.fieldName}}}...{{/each}}`.
+            per_page statements render against the whole batch (`batch_of_records`) rather than a single record — loop with `{{#each batch_of_records}}...{{record.fieldName}}...{{/each}}`, with the same quoting as `query` (double braces for strings, triple for numbers).
 
             `bulk_load` — Stages the records as files and loads them with the database's native COPY / bulk mechanism for maximum throughput; supports INSERT, upsert via bulkLoad.primaryKeys, and custom merge or ignore-existing logic via bulkLoad.overrideMergeQuery; available on Snowflake, Google BigQuery, Microsoft SQL Server (cloud connections), Azure Synapse and Amazon Redshift (NetSuite Analytics Warehouse has it through the JDBC import). The import form calls it "Use optimized bulk load (recommended for larger imports)" and, beside it, calls bulk_insert "Use batch insert SQL query".
             `bulk_insert` — Batch INSERT via multi-row VALUES clause for efficient bulk data loading; has no upsert or ignore-existing logic and is available for all RDBMS types.
@@ -5348,7 +5358,7 @@ components:
               Prompt: "Import customers, ignore existing based on email"
               ```json
               "queryType": ["per_record"],
-              "query": ["INSERT INTO customers (name, email) VALUES ('{{{record.name}}}', '{{{record.email}}}')"]
+              "query": ["INSERT INTO customers (name, email) VALUES ({{record.name}}, {{record.email}})"]
               ```
 
               ### Pure INSERT (no duplicate checking)
@@ -5362,7 +5372,7 @@ components:
               Prompt: "Update inventory counts"
               ```json
               "queryType": ["per_record"],
-              "query": ["UPDATE inventory SET qty = {{{record.qty}}} WHERE sku = '{{{record.sku}}}'"]
+              "query": ["UPDATE inventory SET qty = {{{record.qty}}} WHERE sku = {{record.sku}}"]
               ```
 
               ### Run Once Per Page
@@ -5370,7 +5380,7 @@ components:
               ```json
               "queryType": ["per_page"]
               ```
-              **per_page uses a different Handlebars context than per_record.** The context is the entire batch (`batch_of_records`), not a single `record.` object. Loop through with `{{#each batch_of_records}}...{{{record.fieldName}}}...{{/each}}`.
+              **per_page uses a different Handlebars context than per_record.** The context is the entire batch (`batch_of_records`), not a single `record.` object. Loop through with `{{#each batch_of_records}}...{{record.fieldName}}...{{/each}}`, quoting as in per_record: double braces for strings, triple braces for numbers.
 
               ### Run Once at Start
               Prompt: "Run this cleanup script before processing"
@@ -8330,8 +8340,8 @@ components:
       description: |-
         Configuration for a `ToolImport`, which executes a reusable Celigo Tool resource as the
         import action. The referenced tool defines its own input contract, processing pipeline, and
-        output; this object binds the tool to the import and optionally overrides the connections the
-        tool uses for this invocation.
+        output; this object binds the tool to the import, along with the connections the tool's steps
+        use for this invocation.
       x-celigo-ai-guidance:
         - |-
           A ToolImport is the **wrapper** that lets a flow / API / another Tool consume a
@@ -8382,7 +8392,8 @@ components:
             Per-import overrides applied to the referenced tool at execution time. Tools use a
             bring-your-own-keys connection model: the tool declares abstract connection slots, and the
             consumer binds concrete connections here so one tool definition can run against different
-            systems without being modified. Omit to run the tool with its default connections.
+            systems without being modified. Bind every connection the tool's steps use; to keep the
+            connection a step was authored against, bind it to itself.
           x-celigo-ai-guidance:
             - |-
               Per-consumer overrides applied when the wrapper invokes the underlying Tool.
@@ -8390,10 +8401,10 @@ components:
               imports each reference a `_connectionId`,
               and consumers bind real account connections to those abstract requirements here.
             - |-
-              Omit this block (or omit `connections` inside it) when the wrapper inherits the
-              Tool's authored connections as-is.
-              That's the right default when the Tool was authored against connections the
-              consumer also wants to use.
+              The save accepts a wrapper without bindings, but when the Tool's steps use connections
+              its run then fails with `connection_not_overridden`. Write one `connections` entry per
+              connection the Tool's steps use, even when the wrapper keeps the Tool's authored
+              connection (`_id` equal to `_abstractId`); a Tool whose steps use no connection needs none.
           properties:
             connections:
               type: array
@@ -9187,7 +9198,8 @@ components:
     AiAgentConfig:
       type: object
       description: |
-        AI Agent configuration used by both AiAgentImport and GuardrailImport (ai_agent type).
+        AI agent configuration for an AiAgentImport. An `ai_agent` guardrail is configured by
+        the guardrail spec's `GuardrailAiAgentConfig` instead, which has no `tools` arrays.
 
         Configures which AI provider and model to use, along with instructions, parameter
         tuning, output format, and available tools. Providers come in two families with
@@ -9460,7 +9472,6 @@ components:
                 - gpt-5-mini
                 - gpt-5-nano
                 - gpt-4.1-mini
-                - gpt-4o
               x-celigo-ui-override: Required by the AI agent configuration form. Encoded here to mirror the form so builders produce connectable configurations.
               x-celigo-ai-guidance:
                 - |-
@@ -9540,7 +9551,9 @@ components:
               minimum: 100
               maximum: 128000
               default: 5000
-              description: Maximum number of tokens in the model's response (server default observed live on create)
+              description: |-
+                Caps the length of the model's response, in tokens. An agent created without it is
+                saved with 5000; raise it for long outputs such as large extractions.
             serviceTier:
               type: string
               enum:
@@ -16185,15 +16198,16 @@ components:
             - filter
           description: |-
             Filter applied to incoming records before they are sent to the destination system.
-            Records that match continue through the import; records that don't are silently
-            dropped. Filter expressions reference the incoming record's fields.
+            Records that match continue through the import; records that don't skip this import,
+            counted as ignored, and continue to the next step. Filter expressions reference the
+            incoming record's fields.
           allOf:
             - $ref: '#/components/schemas/Filter'
           x-celigo-ai-guidance:
             - |-
               ## IMPORT-SPECIFIC BEHAVIOR
 
-              **Pre-Import Filtering**: Filters are applied to incoming records before they are sent to the destination system. Records that don't match the filter criteria are silently dropped and not imported.
+              **Pre-Import Filtering**: Filters are applied to incoming records before they are sent to the destination system. A record that doesn't match is not sent, but it is not removed from the flow: it continues to the next step without this import's response-mapped fields, so a later step that writes those fields needs its own filter.
         traceKeyTemplate:
           type: string
           x-celigo-agent:
@@ -16310,7 +16324,7 @@ components:
                   mapping step, on the raw upstream records.  Engage when the
                   user needs to reshape, validate, filter, or annotate records
                   before the mapping sees them (drop archived records,
-                  normalize date strings, split a record into many).  Don't
+                  normalize date strings).  Don't
                   engage for static value substitution or simple field
                   renames — mappings handle those.
 
@@ -16586,7 +16600,7 @@ components:
                 resultField, ...)
               - ``netsuite_da.lookups`` — NetSuite search (recordType,
                 searchField, operator, ...)
-              - ``http.lookups`` — HTTP request (method, relativeURI, postBody,
+              - ``http.lookups`` — HTTP request (method, relativeURI, body,
                 extract, ...)
               - ``jdbc.lookups`` / ``rdbms.lookups`` — SQL query (query,
                 extract, ...)

package/dist/tool.yml

@@ -12883,7 +12883,7 @@ components:
               ## EXAMPLES:
               - `{ "UserId": "12345", "Name": "John Doe", "Age": 30, "Preferences": { "Language": "en", "Notifications": true } }`
           examples:
-            - '{"id": "{{record.id}}", "name": "{{record.name}}", "total": {{record.total}}}'
+            - '{"id": "{{record.id}}", "name": "{{record.name}}", "total": {{record.total}} }'
         updateExpression:
           type: string
           x-celigo-ui-override: Required by the DynamoDB import form when method is updateItem (dynamodb.updateExpression required, visibleWhen method=updateItem).
@@ -13546,7 +13546,10 @@ components:
               the Import's mapping step runs first and shapes the source record,
               then this template renders against that mapped record.
               Reference the mapped record with the ``record`` root when wrapping it (e.g.
-              ``{"requestData": {{{jsonSerialize record}}}}``).
+              ``{"requestData": {{{jsonSerialize record}}} }``).
+              Keep a literal ``}`` apart from a closing ``}}`` or ``}}}`` (a space is enough): the save
+              accepts ``{{{jsonSerialize record}}}}``, but every record then fails with
+              ``handlebars_template_parse_error``.
               Avoid bare ``{{{jsonSerialize record}}}`` with no wrapper — leave the field undefined instead.
               Also avoid wrapping a single post-mapped scalar field in ``jsonSerialize`` (e.g.
               ``{{{jsonSerialize record.<scalar>}}}``) — that pattern is brittle.
@@ -13609,9 +13612,9 @@ components:
 
               Typical shapes:
               - Message-bus envelope:
-                ``{"MessageBody": {"class": "EntityUpdate", "payload": {{{jsonSerialize record}}}}}``
+                ``{"MessageBody": {"class": "EntityUpdate", "payload": {{{jsonSerialize record}}} }}``
               - Metadata wrap:
-                ``{"batchNumber": "{{timestamp dateTimeFormat CST}}", "interfaceType": "...", "requestData": {{{jsonSerialize record}}}}``
+                ``{"batchNumber": "{{timestamp dateTimeFormat CST}}", "interfaceType": "...", "requestData": {{{jsonSerialize record}}} }``
               - Named single-record array wrap (when destination wants ``[record]`` but each request sends one):
                 ``{"products": [{{{jsonSerialize record}}}]}``
 
@@ -13652,7 +13655,7 @@ components:
               current ``{{#each data}}`` item.
 
               Example shape:
-              ``{"header": {"sellerId": "{{connection.http.unencrypted.sellerId}}"}, "messages": [{{#each data}}{"sku": "{{sku}}", "qty": {{quantity}}}{{/each}}]}``
+              ``{"header": {"sellerId": "{{connection.http.unencrypted.sellerId}}"}, "messages": [{{#each data}}{"sku": "{{sku}}", "qty": {{quantity}} }{{/each}}]}``
             - |-
               ## Anti-pattern — DO NOT wrap a post-mapped scalar in jsonSerialize
 
@@ -13737,7 +13740,7 @@ components:
                 mix the two shapes in one body.
           examples:
             - - '{"name": "{{record.name}}", "email": "{{record.email}}"}'
-            - - '{"requestData": {{{jsonSerialize record}}}}'
+            - - '{"requestData": {{{jsonSerialize record}}} }'
             - - '{"values": [{{#each batch_of_records}}{{{jsonSerialize this.record}}}{{#unless @last}},{{/unless}}{{/each}}]}'
           x-celigo-agent:
             - handlebars
@@ -14227,9 +14230,7 @@ components:
 
             The referenced helper bundles the polling config plus a required status
             export (polled to check progress) and a result export (fetches the
-            final payload). An import configured with an async helper cannot carry
-            its own transform, output filter, or preSavePage hook — put that
-            processing in the result export instead.
+            final payload).
           examples:
             - 60a2c4e6f321d800129a1a3c
           x-celigo-agent:
@@ -14313,7 +14314,7 @@ components:
 
               - **Static** — only ``map`` + optional ``default``.  No HTTP
                 request issued.  Use for small controlled vocabularies.
-              - **Dynamic** — ``method`` + ``relativeURI`` (+ ``postBody``,
+              - **Dynamic** — ``method`` + ``relativeURI`` (+ ``body``,
                 ``headers``, ``extract``).  Issues a live HTTP request per
                 lookup.  Adds one HTTP request per record at runtime — factor
                 this into flow design for high-volume syncs.
@@ -14351,14 +14352,18 @@ components:
                   inject values from the incoming record — e.g.
                   ``/customers?email={{{email}}}`` or
                   ``/accounts/{{{accountId}}}``.
-              postBody:
+              body:
                 type: string
                 x-celigo-agent:
                   - handlebars
                 description: |
-                  Request body template for POST / PUT / PATCH lookup methods.
-                  Typically a JSON string with Handlebars placeholders —
-                  ``{"email":"{{{email}}}"}``. Ignored for GET / DELETE.
+                  Request body template for POST / PUT / PATCH lookup methods, typically
+                  a JSON string with Handlebars placeholders (``{"email":"{{{email}}}"}``);
+                  ignored for GET / DELETE. The record's fields sit at the top level:
+                  ``{{email}}`` renders the record's value, while ``{{record.email}}``
+                  renders empty.
+                x-celigo-ai-guidance:
+                  - 'Put the lookup''s request template here, not in ``postBody``: a ``postBody`` saves without an error but is not kept under ``http.lookups``, so the lookup goes out with no body and the run reports no error.'
               headers:
                 type: array
                 items:
@@ -16009,9 +16014,9 @@ components:
                                 line) to the rows that populate the
                                 subrecord.  Typically `$` when the
                                 subrecord data is carried on the same
-                                line item (the enclosing
-                                `lists[].jsonPath` has already selected
-                                the line).  Defaults to `$`.
+                                line item (iteration over the sublist
+                                has already selected the line).
+                                Defaults to `$`.
                             mapping:
                               type: object
                               description: |
@@ -16466,9 +16471,10 @@ components:
                   SQL `SELECT` statement that resolves the lookup value.
                   Typically returns a single row with a single column; use
                   `extract` to pick the column when the query returns multiple.
-                  Inject values from the incoming record with Handlebars — RDBMS
-                  queries use triple braces and the `record.` prefix:
-                  `SELECT id FROM users WHERE email = '{{{record.email}}}'`.
+                  Inject values from the incoming record with Handlebars and the
+                  `record.` prefix — double braces for strings, which the platform
+                  quotes and escapes, and triple braces for numbers:
+                  `SELECT id FROM users WHERE email = {{record.email}}`.
               extract:
                 type: string
                 description: |
@@ -16516,7 +16522,7 @@ components:
           description: |-
             SQL statements to execute; required when queryType is `["per_record"]`, `["per_page"]`, or `["first_page"]`, and typically contains a single statement.
             Each element is a complete SQL statement as a plain string — never a single bare string and never an array of objects (objects cause a Cast error).
-            Inject values from the incoming record with Handlebars using the `record.` prefix and triple braces, e.g. `'{{{record.name}}}'` for strings (single-quoted in the SQL) and `{{{record.quantity}}}` for numbers.
+            Inject values from the incoming record with Handlebars using the `record.` prefix: double braces for strings and dates (`{{record.name}}`; the platform quotes and escapes the value) and triple braces for numbers (`{{{record.quantity}}}`).
             Rendering is strict: a reference the render model cannot resolve (a bare `{{name}}` without prefix, or `record.*` when the arriving record is a grouped array) fails the step with `cannot_evaluate_handlebars` — it does not render an empty value.
             `record` and `rows` are aliases chosen by the arriving data's shape: an object record binds `record` (`rows` undefined); a grouped array-record (e.g. upstream `groupByFields`) binds `rows` (`record` undefined) — address grouped rows as `{{rows.0.field}}` or iterate `{{#each rows}}`.
             Double braces (`{{record.field}}`) output the value wrapped in single quotes with embedded quotes doubled, while triple braces output it raw; block helpers such as `{{#each}}` and `{{#if}}` use double braces as normal.
@@ -16530,7 +16536,7 @@ components:
               - CORRECT: ["INSERT INTO users (name, email) VALUES ({{record.name}}, {{record.email}})"]
               - WRONG: "INSERT INTO users ..."  (not an array)
               - WRONG: [{"query": "INSERT INTO users ..."}]  (objects are invalid — causes Cast error)
-              - WRONG: ["INSERT INTO users (name) VALUES ('{{name}}')"]  (missing record. prefix — rendering is strict, so the step fails with cannot_evaluate_handlebars)
+              - WRONG: ["INSERT INTO users (name) VALUES ({{name}})"]  (missing record. prefix — rendering is strict, so the step fails with cannot_evaluate_handlebars)
             - |-
               ## RDBMS HANDLEBARS SYNTAX
 
@@ -16564,7 +16570,7 @@ components:
                 | Numeric column value (INT, DECIMAL, ...) | `{{{record.field}}}`    |
                 | Stored procedure raw arg (already typed) | `{{{record.field}}}`    |
                 | Identifier (table/column name)           | `{{{record.field}}}`    |
-                | JSON literal (`PARSE_JSON`/JSONB)      | `{{{jsonSerialize record.subobject}}}` |
+                | JSON literal (`PARSE_JSON`/JSONB)      | `{{jsonSerialize record.subobject}}` (renders the JSON as a quoted string) |
 
               - **Block helpers** (`{{#each}}`, `{{#if}}`, `{{/each}}`,
                 `{{#unless}}`) use double braces as normal — they emit no value.
@@ -16583,7 +16589,7 @@ components:
               - Per-page batch (different Handlebars context — see `queryType: per_page` below): `["{{#each batch_of_records}}UPDATE staging SET done=1 WHERE id={{{this.record.id}}};{{/each}}"]`
           examples:
             - - INSERT INTO orders (id, total) VALUES ({{{record.id}}}, {{{record.total}}})
-            - - '{{#each batch_of_records}}UPDATE customers SET name=''{{{record.name}}}'' WHERE id={{{record.id}}};{{/each}}'
+            - - '{{#each batch_of_records}}UPDATE customers SET name={{record.name}} WHERE id={{{record.id}}};{{/each}}'
         queryType:
           type:
             - array
@@ -16612,7 +16618,7 @@ components:
             Prefer the highest-performance type the operation supports: bulk_load (Snowflake, Google BigQuery, Microsoft SQL Server on a cloud connection, Azure Synapse and Amazon Redshift — handles INSERT, upsert via `bulkLoad.primaryKeys`, and custom merge or ignore-existing logic via `bulkLoad.overrideMergeQuery`), then bulk_insert for pure INSERTs, then per_page, then per_record only when the logic cannot be expressed as a batch operation.
             For UPDATE, UPSERT, or ignore-existing logic on databases without bulk_load support, use per_page or per_record with `query` — bulk_insert has no duplicate-checking logic.
             Set a single value; do not combine types or use the legacy INSERT and UPDATE values.
-            per_page statements render against the whole batch (`batch_of_records`) rather than a single record — loop with `{{#each batch_of_records}}...{{{record.fieldName}}}...{{/each}}`.
+            per_page statements render against the whole batch (`batch_of_records`) rather than a single record — loop with `{{#each batch_of_records}}...{{record.fieldName}}...{{/each}}`, with the same quoting as `query` (double braces for strings, triple for numbers).
 
             `bulk_load` — Stages the records as files and loads them with the database's native COPY / bulk mechanism for maximum throughput; supports INSERT, upsert via bulkLoad.primaryKeys, and custom merge or ignore-existing logic via bulkLoad.overrideMergeQuery; available on Snowflake, Google BigQuery, Microsoft SQL Server (cloud connections), Azure Synapse and Amazon Redshift (NetSuite Analytics Warehouse has it through the JDBC import). The import form calls it "Use optimized bulk load (recommended for larger imports)" and, beside it, calls bulk_insert "Use batch insert SQL query".
             `bulk_insert` — Batch INSERT via multi-row VALUES clause for efficient bulk data loading; has no upsert or ignore-existing logic and is available for all RDBMS types.
@@ -16659,7 +16665,7 @@ components:
               Prompt: "Import customers, ignore existing based on email"
               ```json
               "queryType": ["per_record"],
-              "query": ["INSERT INTO customers (name, email) VALUES ('{{{record.name}}}', '{{{record.email}}}')"]
+              "query": ["INSERT INTO customers (name, email) VALUES ({{record.name}}, {{record.email}})"]
               ```
 
               ### Pure INSERT (no duplicate checking)
@@ -16673,7 +16679,7 @@ components:
               Prompt: "Update inventory counts"
               ```json
               "queryType": ["per_record"],
-              "query": ["UPDATE inventory SET qty = {{{record.qty}}} WHERE sku = '{{{record.sku}}}'"]
+              "query": ["UPDATE inventory SET qty = {{{record.qty}}} WHERE sku = {{record.sku}}"]
               ```
 
               ### Run Once Per Page
@@ -16681,7 +16687,7 @@ components:
               ```json
               "queryType": ["per_page"]
               ```
-              **per_page uses a different Handlebars context than per_record.** The context is the entire batch (`batch_of_records`), not a single `record.` object. Loop through with `{{#each batch_of_records}}...{{{record.fieldName}}}...{{/each}}`.
+              **per_page uses a different Handlebars context than per_record.** The context is the entire batch (`batch_of_records`), not a single `record.` object. Loop through with `{{#each batch_of_records}}...{{record.fieldName}}...{{/each}}`, quoting as in per_record: double braces for strings, triple braces for numbers.
 
               ### Run Once at Start
               Prompt: "Run this cleanup script before processing"
@@ -19641,8 +19647,8 @@ components:
       description: |-
         Configuration for a `ToolImport`, which executes a reusable Celigo Tool resource as the
         import action. The referenced tool defines its own input contract, processing pipeline, and
-        output; this object binds the tool to the import and optionally overrides the connections the
-        tool uses for this invocation.
+        output; this object binds the tool to the import, along with the connections the tool's steps
+        use for this invocation.
       x-celigo-ai-guidance:
         - |-
           A ToolImport is the **wrapper** that lets a flow / API / another Tool consume a
@@ -19693,7 +19699,8 @@ components:
             Per-import overrides applied to the referenced tool at execution time. Tools use a
             bring-your-own-keys connection model: the tool declares abstract connection slots, and the
             consumer binds concrete connections here so one tool definition can run against different
-            systems without being modified. Omit to run the tool with its default connections.
+            systems without being modified. Bind every connection the tool's steps use; to keep the
+            connection a step was authored against, bind it to itself.
           x-celigo-ai-guidance:
             - |-
               Per-consumer overrides applied when the wrapper invokes the underlying Tool.
@@ -19701,10 +19708,10 @@ components:
               imports each reference a `_connectionId`,
               and consumers bind real account connections to those abstract requirements here.
             - |-
-              Omit this block (or omit `connections` inside it) when the wrapper inherits the
-              Tool's authored connections as-is.
-              That's the right default when the Tool was authored against connections the
-              consumer also wants to use.
+              The save accepts a wrapper without bindings, but when the Tool's steps use connections
+              its run then fails with `connection_not_overridden`. Write one `connections` entry per
+              connection the Tool's steps use, even when the wrapper keeps the Tool's authored
+              connection (`_id` equal to `_abstractId`); a Tool whose steps use no connection needs none.
           properties:
             connections:
               type: array
@@ -20498,7 +20505,8 @@ components:
     AiAgentConfig:
       type: object
       description: |
-        AI Agent configuration used by both AiAgentImport and GuardrailImport (ai_agent type).
+        AI agent configuration for an AiAgentImport. An `ai_agent` guardrail is configured by
+        the guardrail spec's `GuardrailAiAgentConfig` instead, which has no `tools` arrays.
 
         Configures which AI provider and model to use, along with instructions, parameter
         tuning, output format, and available tools. Providers come in two families with
@@ -20771,7 +20779,6 @@ components:
                 - gpt-5-mini
                 - gpt-5-nano
                 - gpt-4.1-mini
-                - gpt-4o
               x-celigo-ui-override: Required by the AI agent configuration form. Encoded here to mirror the form so builders produce connectable configurations.
               x-celigo-ai-guidance:
                 - |-
@@ -20851,7 +20858,9 @@ components:
               minimum: 100
               maximum: 128000
               default: 5000
-              description: Maximum number of tokens in the model's response (server default observed live on create)
+              description: |-
+                Caps the length of the model's response, in tokens. An agent created without it is
+                saved with 5000; raise it for long outputs such as large extractions.
             serviceTier:
               type: string
               enum:
@@ -25054,15 +25063,16 @@ components:
             - filter
           description: |-
             Filter applied to incoming records before they are sent to the destination system.
-            Records that match continue through the import; records that don't are silently
-            dropped. Filter expressions reference the incoming record's fields.
+            Records that match continue through the import; records that don't skip this import,
+            counted as ignored, and continue to the next step. Filter expressions reference the
+            incoming record's fields.
           allOf:
             - $ref: '#/components/schemas/Filter'
           x-celigo-ai-guidance:
             - |-
               ## IMPORT-SPECIFIC BEHAVIOR
 
-              **Pre-Import Filtering**: Filters are applied to incoming records before they are sent to the destination system. Records that don't match the filter criteria are silently dropped and not imported.
+              **Pre-Import Filtering**: Filters are applied to incoming records before they are sent to the destination system. A record that doesn't match is not sent, but it is not removed from the flow: it continues to the next step without this import's response-mapped fields, so a later step that writes those fields needs its own filter.
         traceKeyTemplate:
           type: string
           x-celigo-agent:
@@ -25179,7 +25189,7 @@ components:
                   mapping step, on the raw upstream records.  Engage when the
                   user needs to reshape, validate, filter, or annotate records
                   before the mapping sees them (drop archived records,
-                  normalize date strings, split a record into many).  Don't
+                  normalize date strings).  Don't
                   engage for static value substitution or simple field
                   renames — mappings handle those.
 
@@ -25455,7 +25465,7 @@ components:
                 resultField, ...)
               - ``netsuite_da.lookups`` — NetSuite search (recordType,
                 searchField, operator, ...)
-              - ``http.lookups`` — HTTP request (method, relativeURI, postBody,
+              - ``http.lookups`` — HTTP request (method, relativeURI, body,
                 extract, ...)
               - ``jdbc.lookups`` / ``rdbms.lookups`` — SQL query (query,
                 extract, ...)
@@ -30514,7 +30524,7 @@ components:
               (`http.paging.body` likewise from its paging fields). A body that is not parseable JSON at
               rest — Handlebars block helpers outside a string value — renders every field EMPTY in that
               form, so keep templated variables inside a JSON string value:
-              `"variables": "{\"after\": {{#if export.http.paging.token}}\"{{export.http.paging.token}}\"{{else}}null{{/if}}}"`.
+              `"variables": "{\"after\": {{#if export.http.paging.token}}\"{{export.http.paging.token}}\"{{else}}null{{/if}} }"`.
             - |-
               The GraphQL form has no relative URI field: the connection `baseURI` is expected to be the
               GraphQL endpoint (a saved `relativeURI` still applies but is not shown). Create GraphQL

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@celigo/api-specs",
-  "version": "0.2.131",
+  "version": "0.2.133",
   "private": false,
   "description": "Public OpenAPI bundles (dist/) and the celigo-cli lint schema pack (schemas.json).",
   "keywords": [

package/schemas.json

@@ -1,7 +1,7 @@
 {
 	"provenance": {
 		"repo": "celigo/integrator-api-specs",
-		"iasCommit": "a1fc60f9",
+		"iasCommit": "60c8d128",
 		"generatedAt": "2026-10-21"
 	},
 	"roots": {
@@ -6674,13 +6674,13 @@
 			"p": {
 				"_lookupCacheId": 1,
 				"allowFailures": 118,
+				"body": 1,
 				"default": 88,
 				"extract": 1,
 				"headers": 260,
 				"map": 117,
 				"method": 164,
 				"name": 1,
-				"postBody": 1,
 				"relativeURI": 1,
 				"useDefaultOnMultipleMatches": 9
 			}