drupal-mcp-connector 2.7.3 → 2.8.0

package/.agents/commands/drupal-audit-config-best-practices.md

@@ -0,0 +1,15 @@
+---
+description: "Lint key configuration for production-readiness and security: on-screen error display, CSS/JS aggregation, anonymous page cache, open user registration, missing 404/403 pages, and insecure file uploads. Severity-ranked. Requires config-read access and the server-tool bridge."
+argument-hint: "[site]"
+---
+
+Call the MCP tool `drupal_audit_config_best_practices`.
+
+Lint key configuration for production-readiness and security: on-screen error display, CSS/JS aggregation, anonymous page cache, open user registration, missing 404/403 pages, and insecure file uploads. Severity-ranked. Requires config-read access and the server-tool bridge.
+
+Parse the arguments supplied with this command into this tool's parameters:
+
+**Optional:**
+- `site` (string): Named site from connector config. Omit only on reads: multi-site configs fall back to defaultSite (often local/dev, not production). Writes require an explicit site when more than one site is configured. Every response includes `_target` { name, baseUrl, source } (`hint` when you passed site, `default` when you did not).
+
+If a required parameter is missing, ask before calling — do not invent values. Coerce each value to its JSON type (booleans → true/false, numbers → numeric, object/array → parse JSON), then make the single tool call and summarize the result.

package/{.claude → .agents}/commands/drupal-audit-site-health.md

@@ -1,19 +1,18 @@
 ---
 description: "Composite site-health dashboard: runs a battery of content, link, and configuration audits and rolls them into one scored report with a letter grade. Each section degrades independently — gated/errored sections are recorded, not fatal. Read-only."
 argument-hint: "[site] [type] [sampleSize] [sections]"
-allowed-tools: mcp__drupal__drupal_audit_site_health
 ---
 
-Call the `mcp__drupal__drupal_audit_site_health` MCP tool.
+Call the MCP tool `drupal_audit_site_health`.
 
 Composite site-health dashboard: runs a battery of content, link, and configuration audits and rolls them into one scored report with a letter grade. Each section degrades independently — gated/errored sections are recorded, not fatal. Read-only.
 
-Parse the request in `$ARGUMENTS` into this tool's parameters:
+Parse the arguments supplied with this command into this tool's parameters:
 
 **Optional:**
-- `site` (string): omit for the default site
+- `site` (string): Named site from connector config. Omit only on reads: multi-site configs fall back to defaultSite (often local/dev, not production). Writes require an explicit site when more than one site is configured. Every response includes `_target` { name, baseUrl, source } (`hint` when you passed site, `default` when you did not).
 - `type` (string): Primary content type for content/link sections (default: article)
 - `sampleSize` (number): Per-section scan cap (kept small for a fast roll-up)
 - `sections` (array (pass as JSON)): Subset of sections to run (default: all). Available: content_summary, stale_content, seo_audit, accessibility, duplicate_content, readability, pii_exposure, broken_links, redirect_health, alias_coverage, log_404, config_best_practices, module_audit, permission_audit
 
-If a required parameter is missing from `$ARGUMENTS`, ask before calling — do not invent values. Coerce each value to its JSON type (booleans → true/false, numbers → numeric, object/array → parse JSON), then make the single tool call and summarize the result.
+If a required parameter is missing, ask before calling — do not invent values. Coerce each value to its JSON type (booleans → true/false, numbers → numeric, object/array → parse JSON), then make the single tool call and summarize the result.

package/.agents/commands/drupal-block-user.md

@@ -0,0 +1,19 @@
+---
+description: "Block or unblock a Drupal user account without deleting it."
+argument-hint: "<id> [site] [block]"
+---
+
+Call the MCP tool `drupal_block_user`.
+
+Block or unblock a Drupal user account without deleting it.
+
+Parse the arguments supplied with this command into this tool's parameters:
+
+**Required:**
+- `id` (string): User UUID
+
+**Optional:**
+- `site` (string): Named site from connector config. Omit only on reads: multi-site configs fall back to defaultSite (often local/dev, not production). Writes require an explicit site when more than one site is configured. Every response includes `_target` { name, baseUrl, source } (`hint` when you passed site, `default` when you did not).
+- `block` (boolean (true/false)): true = block, false = unblock
+
+If a required parameter is missing, ask before calling — do not invent values. Coerce each value to its JSON type (booleans → true/false, numbers → numeric, object/array → parse JSON), then make the single tool call and summarize the result.

package/{.claude → .agents}/commands/drupal-bulk-create.md

@@ -1,14 +1,13 @@
 ---
 description: "Create many entities of a single type + bundle in one call. Permission is checked once; each item is created independently, so the batch continues past individual failures (partial success). Returns per-item { index, success, id | error } and a summary { created, failed }. Paragraph items also return relationshipData with meta.target_revision_id for a later host attach. Writes default to unpublished/draft."
 argument-hint: "<entityType> <bundle> <items> [site]"
-allowed-tools: mcp__drupal__drupal_bulk_create
 ---
 
-Call the `mcp__drupal__drupal_bulk_create` MCP tool.
+Call the MCP tool `drupal_bulk_create`.
 
 Create many entities of a single type + bundle in one call. Permission is checked once; each item is created independently, so the batch continues past individual failures (partial success). Returns per-item { index, success, id | error } and a summary { created, failed }. Paragraph items also return relationshipData with meta.target_revision_id for a later host attach. Writes default to unpublished/draft.
 
-Parse the request in `$ARGUMENTS` into this tool's parameters:
+Parse the arguments supplied with this command into this tool's parameters:
 
 **Required:**
 - `entityType` (string): Entity type machine name, e.g. 'node', 'taxonomy_term'
@@ -16,6 +15,6 @@ Parse the request in `$ARGUMENTS` into this tool's parameters:
 - `items` (array (pass as JSON)): Entities to create. Each is { attributes?, relationships? }.
 
 **Optional:**
-- `site` (string): omit for the default site
+- `site` (string): Named site from connector config. Omit only on reads: multi-site configs fall back to defaultSite (often local/dev, not production). Writes require an explicit site when more than one site is configured. Every response includes `_target` { name, baseUrl, source } (`hint` when you passed site, `default` when you did not).
 
-If a required parameter is missing from `$ARGUMENTS`, ask before calling — do not invent values. Coerce each value to its JSON type (booleans → true/false, numbers → numeric, object/array → parse JSON), then make the single tool call and summarize the result.
+If a required parameter is missing, ask before calling — do not invent values. Coerce each value to its JSON type (booleans → true/false, numbers → numeric, object/array → parse JSON), then make the single tool call and summarize the result.

package/{.claude → .agents}/commands/drupal-bulk-update.md

@@ -1,14 +1,13 @@
 ---
 description: "Update many entities of a single type + bundle in one call. Permission is checked once; each item is updated independently, so the batch continues past individual failures (partial success). Each item requires an 'id' (UUID); items missing an id are reported as per-item failures. Published moderated targets without an explicit attributes.moderation_state default to moderation_state 'draft' (forward revision) so bulk edits do not mutate live default revisions. Returns per-item { index, success, id | error } and a summary { updated, failed }."
 argument-hint: "<entityType> <bundle> <items> [site]"
-allowed-tools: mcp__drupal__drupal_bulk_update
 ---
 
-Call the `mcp__drupal__drupal_bulk_update` MCP tool.
+Call the MCP tool `drupal_bulk_update`.
 
 Update many entities of a single type + bundle in one call. Permission is checked once; each item is updated independently, so the batch continues past individual failures (partial success). Each item requires an 'id' (UUID); items missing an id are reported as per-item failures. Published moderated targets without an explicit attributes.moderation_state default to moderation_state 'draft' (forward revision) so bulk edits do not mutate live default revisions. Returns per-item { index, success, id | error } and a summary { updated, failed }.
 
-Parse the request in `$ARGUMENTS` into this tool's parameters:
+Parse the arguments supplied with this command into this tool's parameters:
 
 **Required:**
 - `entityType` (string): Entity type machine name
@@ -16,6 +15,6 @@ Parse the request in `$ARGUMENTS` into this tool's parameters:
 - `items` (array (pass as JSON)): Entities to update. Each is { id, attributes?, relationships? }.
 
 **Optional:**
-- `site` (string): omit for the default site
+- `site` (string): Named site from connector config. Omit only on reads: multi-site configs fall back to defaultSite (often local/dev, not production). Writes require an explicit site when more than one site is configured. Every response includes `_target` { name, baseUrl, source } (`hint` when you passed site, `default` when you did not).
 
-If a required parameter is missing from `$ARGUMENTS`, ask before calling — do not invent values. Coerce each value to its JSON type (booleans → true/false, numbers → numeric, object/array → parse JSON), then make the single tool call and summarize the result.
+If a required parameter is missing, ask before calling — do not invent values. Coerce each value to its JSON type (booleans → true/false, numbers → numeric, object/array → parse JSON), then make the single tool call and summarize the result.

package/.agents/commands/drupal-config-get.md

@@ -0,0 +1,18 @@
+---
+description: "Read a single Drupal configuration object by name (e.g. \"system.site\") via the governed server-side config tool. Requires config read access."
+argument-hint: "<name> [site]"
+---
+
+Call the MCP tool `drupal_config_get`.
+
+Read a single Drupal configuration object by name (e.g. "system.site") via the governed server-side config tool. Requires config read access.
+
+Parse the arguments supplied with this command into this tool's parameters:
+
+**Required:**
+- `name` (string)
+
+**Optional:**
+- `site` (string): Named site from connector config. Omit only on reads: multi-site configs fall back to defaultSite (often local/dev, not production). Writes require an explicit site when more than one site is configured. Every response includes `_target` { name, baseUrl, source } (`hint` when you passed site, `default` when you did not).
+
+If a required parameter is missing, ask before calling — do not invent values. Coerce each value to its JSON type (booleans → true/false, numbers → numeric, object/array → parse JSON), then make the single tool call and summarize the result.

package/.agents/commands/drupal-config-list.md

@@ -0,0 +1,16 @@
+---
+description: "List Drupal configuration object names, optionally filtered by a name prefix, via the governed server-side config tool. Requires config read access."
+argument-hint: "[site] [prefix]"
+---
+
+Call the MCP tool `drupal_config_list`.
+
+List Drupal configuration object names, optionally filtered by a name prefix, via the governed server-side config tool. Requires config read access.
+
+Parse the arguments supplied with this command into this tool's parameters:
+
+**Optional:**
+- `site` (string): Named site from connector config. Omit only on reads: multi-site configs fall back to defaultSite (often local/dev, not production). Writes require an explicit site when more than one site is configured. Every response includes `_target` { name, baseUrl, source } (`hint` when you passed site, `default` when you did not).
+- `prefix` (string)
+
+If a required parameter is missing, ask before calling — do not invent values. Coerce each value to its JSON type (booleans → true/false, numbers → numeric, object/array → parse JSON), then make the single tool call and summarize the result.

package/.agents/commands/drupal-config-set.md

@@ -0,0 +1,19 @@
+---
+description: "Set a Drupal configuration value via the governed server-side config tool. Audited and gated server-side; requires the config-editor (Developer) tier. Then export to YAML for a PR."
+argument-hint: "<name> <value> [site]"
+---
+
+Call the MCP tool `drupal_config_set`.
+
+Set a Drupal configuration value via the governed server-side config tool. Audited and gated server-side; requires the config-editor (Developer) tier. Then export to YAML for a PR.
+
+Parse the arguments supplied with this command into this tool's parameters:
+
+**Required:**
+- `name` (string)
+- `value` (object (pass as JSON)): A map of top-level config keys to their new values (e.g. { "slogan": "Information Technology" }). Other keys in the object are preserved server-side.
+
+**Optional:**
+- `site` (string): Named site from connector config. Omit only on reads: multi-site configs fall back to defaultSite (often local/dev, not production). Writes require an explicit site when more than one site is configured. Every response includes `_target` { name, baseUrl, source } (`hint` when you passed site, `default` when you did not).
+
+If a required parameter is missing, ask before calling — do not invent values. Coerce each value to its JSON type (booleans → true/false, numbers → numeric, object/array → parse JSON), then make the single tool call and summarize the result.

package/.agents/commands/drupal-content-by-moderation-state.md

@@ -0,0 +1,21 @@
+---
+description: "List nodes of a content type currently in a given moderation state (e.g. what is in 'draft' or 'needs_review'). Stock JSON:API cannot filter the computed moderation_state field; when the site rejects that filter the tool samples recent nodes client-side and marks the result approximate, instead of returning Drupal's 500."
+argument-hint: "<type> <state> [site] [limit] [offset]"
+---
+
+Call the MCP tool `drupal_content_by_moderation_state`.
+
+List nodes of a content type currently in a given moderation state (e.g. what is in 'draft' or 'needs_review'). Stock JSON:API cannot filter the computed moderation_state field; when the site rejects that filter the tool samples recent nodes client-side and marks the result approximate, instead of returning Drupal's 500.
+
+Parse the arguments supplied with this command into this tool's parameters:
+
+**Required:**
+- `type` (string): Content type machine name
+- `state` (string): Moderation state machine name
+
+**Optional:**
+- `site` (string): Named site from connector config. Omit only on reads: multi-site configs fall back to defaultSite (often local/dev, not production). Writes require an explicit site when more than one site is configured. Every response includes `_target` { name, baseUrl, source } (`hint` when you passed site, `default` when you did not).
+- `limit` (number)
+- `offset` (number)
+
+If a required parameter is missing, ask before calling — do not invent values. Coerce each value to its JSON type (booleans → true/false, numbers → numeric, object/array → parse JSON), then make the single tool call and summarize the result.

package/.agents/commands/drupal-create-block.md

@@ -0,0 +1,20 @@
+---
+description: "Create a custom content block of a given block type (bundle). 'info' is the administrative label; 'body' is optional rendered HTML. Checked against the site security config."
+argument-hint: "<type> <info> [site] [body]"
+---
+
+Call the MCP tool `drupal_create_block`.
+
+Create a custom content block of a given block type (bundle). 'info' is the administrative label; 'body' is optional rendered HTML. Checked against the site security config.
+
+Parse the arguments supplied with this command into this tool's parameters:
+
+**Required:**
+- `type` (string): Block type (bundle) machine name, e.g. 'basic'
+- `info` (string): Administrative label / block description
+
+**Optional:**
+- `site` (string): Named site from connector config. Omit only on reads: multi-site configs fall back to defaultSite (often local/dev, not production). Writes require an explicit site when more than one site is configured. Every response includes `_target` { name, baseUrl, source } (`hint` when you passed site, `default` when you did not).
+- `body` (string): Block body HTML (optional)
+
+If a required parameter is missing, ask before calling — do not invent values. Coerce each value to its JSON type (booleans → true/false, numbers → numeric, object/array → parse JSON), then make the single tool call and summarize the result.

package/{.claude → .agents}/commands/drupal-create-media.md

@@ -1,22 +1,21 @@
 ---
 description: "Create a media entity. For remote video (YouTube/Vimeo), pass the URL via fields.field_media_oembed_video. For file-based media, use drupal_upload_file first to get a file UUID, then pass it in fields."
 argument-hint: "<type> <name> [site] [status] [fields]"
-allowed-tools: mcp__drupal__drupal_create_media
 ---
 
-Call the `mcp__drupal__drupal_create_media` MCP tool.
+Call the MCP tool `drupal_create_media`.
 
 Create a media entity. For remote video (YouTube/Vimeo), pass the URL via fields.field_media_oembed_video. For file-based media, use drupal_upload_file first to get a file UUID, then pass it in fields.
 
-Parse the request in `$ARGUMENTS` into this tool's parameters:
+Parse the arguments supplied with this command into this tool's parameters:
 
 **Required:**
 - `type` (string): Media type machine name
 - `name` (string): Media entity name / label
 
 **Optional:**
-- `site` (string): omit for the default site
+- `site` (string): Named site from connector config. Omit only on reads: multi-site configs fall back to defaultSite (often local/dev, not production). Writes require an explicit site when more than one site is configured. Every response includes `_target` { name, baseUrl, source } (`hint` when you passed site, `default` when you did not).
 - `status` (boolean (true/false)): Published flag. Defaults to false (unpublished). Requires allowPublish when true.
 - `fields` (object (pass as JSON)): Additional field values — include the source field (e.g. field_media_oembed_video: 'https://youtu.be/...'). Entity-reference values in JSON:API linkage shape ({ data: { type, id } }) are sent as relationships automatically.
 
-If a required parameter is missing from `$ARGUMENTS`, ask before calling — do not invent values. Coerce each value to its JSON type (booleans → true/false, numbers → numeric, object/array → parse JSON), then make the single tool call and summarize the result.
+If a required parameter is missing, ask before calling — do not invent values. Coerce each value to its JSON type (booleans → true/false, numbers → numeric, object/array → parse JSON), then make the single tool call and summarize the result.

package/{.claude → .agents}/commands/drupal-create-menu-link.md

@@ -1,14 +1,13 @@
 ---
 description: "Create a custom menu link, enabled by default so it renders immediately. The link target is a Drupal URI such as 'internal:/about', 'entity:node/42', or an absolute 'https://…' URL — prefer 'entity:node/<id>' when linking to a node (avoids the alias-resolution 'path inaccessible' race). Set 'parent' (a parent link plugin id like 'menu_link_content:<uuid>') to nest the link, and 'enabled: false' to create it disabled. Checked against the site security config."
 argument-hint: "<title> <link> <menu> [site] [weight] [parent] [enabled]"
-allowed-tools: mcp__drupal__drupal_create_menu_link
 ---
 
-Call the `mcp__drupal__drupal_create_menu_link` MCP tool.
+Call the MCP tool `drupal_create_menu_link`.
 
 Create a custom menu link, enabled by default so it renders immediately. The link target is a Drupal URI such as 'internal:/about', 'entity:node/42', or an absolute 'https://…' URL — prefer 'entity:node/<id>' when linking to a node (avoids the alias-resolution 'path inaccessible' race). Set 'parent' (a parent link plugin id like 'menu_link_content:<uuid>') to nest the link, and 'enabled: false' to create it disabled. Checked against the site security config.
 
-Parse the request in `$ARGUMENTS` into this tool's parameters:
+Parse the arguments supplied with this command into this tool's parameters:
 
 **Required:**
 - `title` (string): Link label shown in the menu
@@ -16,9 +15,9 @@ Parse the request in `$ARGUMENTS` into this tool's parameters:
 - `menu` (string): Menu machine name to place the link in, e.g. 'main' or 'footer'
 
 **Optional:**
-- `site` (string): omit for the default site
+- `site` (string): Named site from connector config. Omit only on reads: multi-site configs fall back to defaultSite (often local/dev, not production). Writes require an explicit site when more than one site is configured. Every response includes `_target` { name, baseUrl, source } (`hint` when you passed site, `default` when you did not).
 - `weight` (number): Ordering weight within the menu (lower sorts first)
 - `parent` (string): Parent link plugin id to nest under, e.g. 'menu_link_content:<uuid>'. Omit for a top-level link.
 - `enabled` (boolean (true/false)): Whether the link is enabled (renders). Defaults to true.
 
-If a required parameter is missing from `$ARGUMENTS`, ask before calling — do not invent values. Coerce each value to its JSON type (booleans → true/false, numbers → numeric, object/array → parse JSON), then make the single tool call and summarize the result.
+If a required parameter is missing, ask before calling — do not invent values. Coerce each value to its JSON type (booleans → true/false, numbers → numeric, object/array → parse JSON), then make the single tool call and summarize the result.

package/{.claude → .agents}/commands/drupal-create-node.md

@@ -1,29 +1,28 @@
 ---
 description: "Create a new content node. Returns the new node UUID, integer ID, and URL. For content types under an editorial (content_moderation) workflow, set moderationState (e.g. 'draft'/'published') instead of status. Entity-reference fields (taxonomy terms, related content, media) go in `relationships`, not `fields`."
 argument-hint: "<type> <title> [site] [body] [summary] [format] [status] [moderationState] [fields] [relationships] [dryRun] [returning]"
-allowed-tools: mcp__drupal__drupal_create_node
 ---
 
-Call the `mcp__drupal__drupal_create_node` MCP tool.
+Call the MCP tool `drupal_create_node`.
 
 Create a new content node. Returns the new node UUID, integer ID, and URL. For content types under an editorial (content_moderation) workflow, set moderationState (e.g. 'draft'/'published') instead of status. Entity-reference fields (taxonomy terms, related content, media) go in `relationships`, not `fields`.
 
-Parse the request in `$ARGUMENTS` into this tool's parameters:
+Parse the arguments supplied with this command into this tool's parameters:
 
 **Required:**
 - `type` (string): Content type machine name
 - `title` (string)
 
 **Optional:**
-- `site` (string): omit for the default site
+- `site` (string): Named site from connector config. Omit only on reads: multi-site configs fall back to defaultSite (often local/dev, not production). Writes require an explicit site when more than one site is configured. Every response includes `_target` { name, baseUrl, source } (`hint` when you passed site, `default` when you did not).
 - `body` (string): Body field HTML
-- `summary` (string): Body summary/teaser — writes the `summary` property of the body field (core `text_with_summary`). Many headless sites instead use a dedicated summary/deck field for teasers and meta descriptions; on those, set that field in `fields` — a value written here will be stored but may never be rendered.
-- `format` (string): Text format machine name for the body, e.g. 'basic_html'. Defaults to the site config's `defaultTextFormat`, then 'full_html'. Set this when the site's formats do not include full_html, or to avoid writing content into a more permissive format than intended.
+- `summary` (string): Body summary/teaser — writes body.summary on core text_with_summary only. Refused when the sampled body field has no summary property (text_long / text_formatted) or the schema cannot be determined. Prefer the site's dedicated deck/summary field via `fields`.
+- `format` (string): Text format machine name for the body, e.g. 'basic_html'. When the body field's allowed_formats lists exactly one format, that is the default. A caller format outside that list is refused before write. When allowed_formats cannot be resolved, defaults to the site config's `defaultTextFormat`, then 'full_html'.
 - `status` (boolean (true/false)): Published flag for NON-moderated types. true to publish immediately. Ignored if moderationState is set; on a moderated type it is dropped automatically.
 - `moderationState` (string): Moderation state for content_moderation types, e.g. 'draft' or 'published'. Takes precedence over status.
-- `fields` (object (pass as JSON)): Scalar/attribute field values keyed by Drupal machine name. Do NOT put entity-reference fields here — Drupal rejects them as attributes; use `relationships`.
+- `fields` (object (pass as JSON)): Scalar/attribute field values keyed by Drupal machine name. Formatted text: a string or { value, format?, summary? }. format must be in the field's allowed_formats; a single allowed format is used when omitted. Do NOT put entity-reference fields here — Drupal rejects them as attributes; use `relationships`.
 - `relationships` (object (pass as JSON)): Entity-reference fields as JSON:API relationships, keyed by field machine name. Single-value: { field_resource_type: { data: { type: 'taxonomy_term--resource_type', id: '<uuid>' } } }. Multi-value: { field_tags: { data: [{ type: 'taxonomy_term--tags', id: '<uuid>' }] } }.
 - `dryRun` (boolean (true/false)): Validate and return a preview of the write without committing.
 - `returning` (string): Response verbosity. "full" (default) returns the complete saved entity; "minimal" returns just identity + state (id, type, bundle, title, status, changed, url) — much smaller, recommended for bulk writes where the echoed body would dominate the response.
 
-If a required parameter is missing from `$ARGUMENTS`, ask before calling — do not invent values. Coerce each value to its JSON type (booleans → true/false, numbers → numeric, object/array → parse JSON), then make the single tool call and summarize the result.
+If a required parameter is missing, ask before calling — do not invent values. Coerce each value to its JSON type (booleans → true/false, numbers → numeric, object/array → parse JSON), then make the single tool call and summarize the result.

package/{.claude → .agents}/commands/drupal-create-paragraph.md

@@ -1,20 +1,19 @@
 ---
 description: "Create a Paragraph entity of a given paragraph type (bundle). Paragraphs are content fragments that are NOT standalone — they must be referenced by a host entity's paragraph / Entity Reference Revisions field. Returns the created paragraph plus `relationshipData` ({ type: 'paragraph--<bundle>', id, meta: { target_revision_id } }) to drop into a host field's relationships via drupal_entity_update / drupal_update_node. Drupal ERR items are empty without that meta key — do not send {type, id} alone. Before creating paragraphs to attach to a published moderated node, call drupal_list_revisions (possiblyPatchBlocked) and dryRun the host update so a doomed PATCH does not orphan them. Use drupal_get_entity_schema (entityType 'paragraph', the bundle) first to discover fields. Governed by the site security policy."
 argument-hint: "<paragraphType> [site] [attributes]"
-allowed-tools: mcp__drupal__drupal_create_paragraph
 ---
 
-Call the `mcp__drupal__drupal_create_paragraph` MCP tool.
+Call the MCP tool `drupal_create_paragraph`.
 
 Create a Paragraph entity of a given paragraph type (bundle). Paragraphs are content fragments that are NOT standalone — they must be referenced by a host entity's paragraph / Entity Reference Revisions field. Returns the created paragraph plus `relationshipData` ({ type: 'paragraph--<bundle>', id, meta: { target_revision_id } }) to drop into a host field's relationships via drupal_entity_update / drupal_update_node. Drupal ERR items are empty without that meta key — do not send {type, id} alone. Before creating paragraphs to attach to a published moderated node, call drupal_list_revisions (possiblyPatchBlocked) and dryRun the host update so a doomed PATCH does not orphan them. Use drupal_get_entity_schema (entityType 'paragraph', the bundle) first to discover fields. Governed by the site security policy.
 
-Parse the request in `$ARGUMENTS` into this tool's parameters:
+Parse the arguments supplied with this command into this tool's parameters:
 
 **Required:**
 - `paragraphType` (string): Paragraph type / bundle machine name, e.g. 'text', 'image', 'cta'
 
 **Optional:**
-- `site` (string): Named site (omit for default)
+- `site` (string): Named site from connector config. Omit only on reads: multi-site configs fall back to defaultSite (often local/dev, not production). Writes require an explicit site when more than one site is configured. Every response includes `_target` { name, baseUrl, source } (`hint` when you passed site, `default` when you did not).
 - `attributes` (object (pass as JSON)): Paragraph field values keyed by Drupal machine name, e.g. { field_body: { value: '<p>..</p>', format: 'full_html' } }
 
-If a required parameter is missing from `$ARGUMENTS`, ask before calling — do not invent values. Coerce each value to its JSON type (booleans → true/false, numbers → numeric, object/array → parse JSON), then make the single tool call and summarize the result.
+If a required parameter is missing, ask before calling — do not invent values. Coerce each value to its JSON type (booleans → true/false, numbers → numeric, object/array → parse JSON), then make the single tool call and summarize the result.

package/{.claude → .agents}/commands/drupal-create-redirect.md

@@ -1,22 +1,21 @@
 ---
 description: "Create an active URL redirect (contrib Redirect module). The redirect serves its 301 (or chosen code) immediately: 'source' is the old path (a leading slash is fine — it is normalized to the module's stored, slash-less form so the redirect actually matches and fires), and 'target' is the destination as a path ('/new'), an 'entity:node/ID', or an absolute URL. status_code defaults to 301; pass 302 for a temporary redirect. Governed by the site security policy (needs redirect write / 'administer redirects')."
 argument-hint: "<source> <target> [site] [statusCode] [language]"
-allowed-tools: mcp__drupal__drupal_create_redirect
 ---
 
-Call the `mcp__drupal__drupal_create_redirect` MCP tool.
+Call the MCP tool `drupal_create_redirect`.
 
 Create an active URL redirect (contrib Redirect module). The redirect serves its 301 (or chosen code) immediately: 'source' is the old path (a leading slash is fine — it is normalized to the module's stored, slash-less form so the redirect actually matches and fires), and 'target' is the destination as a path ('/new'), an 'entity:node/ID', or an absolute URL. status_code defaults to 301; pass 302 for a temporary redirect. Governed by the site security policy (needs redirect write / 'administer redirects').
 
-Parse the request in `$ARGUMENTS` into this tool's parameters:
+Parse the arguments supplied with this command into this tool's parameters:
 
 **Required:**
 - `source` (string): Source/old path to redirect from, e.g. '/old-slug'. Leading slash optional.
 - `target` (string): Destination: a path ('/new-slug'), 'entity:node/42', or an absolute 'https://…' URL.
 
 **Optional:**
-- `site` (string): Named site (omit for default)
+- `site` (string): Named site from connector config. Omit only on reads: multi-site configs fall back to defaultSite (often local/dev, not production). Writes require an explicit site when more than one site is configured. Every response includes `_target` { name, baseUrl, source } (`hint` when you passed site, `default` when you did not).
 - `statusCode` (number): HTTP redirect status code. 301 (permanent, default) or 302 (temporary); 303/307/308 also accepted.
 - `language` (string): Langcode the redirect applies to. Defaults to 'und' (all languages).
 
-If a required parameter is missing from `$ARGUMENTS`, ask before calling — do not invent values. Coerce each value to its JSON type (booleans → true/false, numbers → numeric, object/array → parse JSON), then make the single tool call and summarize the result.
+If a required parameter is missing, ask before calling — do not invent values. Coerce each value to its JSON type (booleans → true/false, numbers → numeric, object/array → parse JSON), then make the single tool call and summarize the result.

package/.agents/commands/drupal-create-taxonomy-term.md

@@ -0,0 +1,22 @@
+---
+description: "Create a new taxonomy term in a vocabulary."
+argument-hint: "<vocabulary> <name> [site] [description] [weight] [parentId]"
+---
+
+Call the MCP tool `drupal_create_taxonomy_term`.
+
+Create a new taxonomy term in a vocabulary.
+
+Parse the arguments supplied with this command into this tool's parameters:
+
+**Required:**
+- `vocabulary` (string)
+- `name` (string)
+
+**Optional:**
+- `site` (string): Named site from connector config. Omit only on reads: multi-site configs fall back to defaultSite (often local/dev, not production). Writes require an explicit site when more than one site is configured. Every response includes `_target` { name, baseUrl, source } (`hint` when you passed site, `default` when you did not).
+- `description` (string)
+- `weight` (number)
+- `parentId` (string): UUID of parent term (for hierarchical vocabularies)
+
+If a required parameter is missing, ask before calling — do not invent values. Coerce each value to its JSON type (booleans → true/false, numbers → numeric, object/array → parse JSON), then make the single tool call and summarize the result.

package/{.claude → .agents}/commands/drupal-create-translation.md

@@ -1,14 +1,13 @@
 ---
 description: "Create or replace a translation of a Drupal entity for a target language (governed write). Sets the given langcode plus the supplied translated field values. Requires the content_translation module enabled and the bundle configured as translatable; otherwise Drupal rejects the write. Defaults to node."
 argument-hint: "<type> <id> <langcode> [site] [entityType] [attributes]"
-allowed-tools: mcp__drupal__drupal_create_translation
 ---
 
-Call the `mcp__drupal__drupal_create_translation` MCP tool.
+Call the MCP tool `drupal_create_translation`.
 
 Create or replace a translation of a Drupal entity for a target language (governed write). Sets the given langcode plus the supplied translated field values. Requires the content_translation module enabled and the bundle configured as translatable; otherwise Drupal rejects the write. Defaults to node.
 
-Parse the request in `$ARGUMENTS` into this tool's parameters:
+Parse the arguments supplied with this command into this tool's parameters:
 
 **Required:**
 - `type` (string): Bundle machine name, e.g. 'article'
@@ -16,8 +15,8 @@ Parse the request in `$ARGUMENTS` into this tool's parameters:
 - `langcode` (string): Target language code, e.g. 'de', 'fr', 'pt_br'
 
 **Optional:**
-- `site` (string): omit for the default site
+- `site` (string): Named site from connector config. Omit only on reads: multi-site configs fall back to defaultSite (often local/dev, not production). Writes require an explicit site when more than one site is configured. Every response includes `_target` { name, baseUrl, source } (`hint` when you passed site, `default` when you did not).
 - `entityType` (string): Entity type machine name. Default: 'node'.
 - `attributes` (object (pass as JSON)): Translated field values keyed by Drupal machine name
 
-If a required parameter is missing from `$ARGUMENTS`, ask before calling — do not invent values. Coerce each value to its JSON type (booleans → true/false, numbers → numeric, object/array → parse JSON), then make the single tool call and summarize the result.
+If a required parameter is missing, ask before calling — do not invent values. Coerce each value to its JSON type (booleans → true/false, numbers → numeric, object/array → parse JSON), then make the single tool call and summarize the result.

package/.agents/commands/drupal-create-user.md

@@ -0,0 +1,23 @@
+---
+description: "Create a new Drupal user account with optional roles and password."
+argument-hint: "<name> <mail> [site] [password] [status] [roles] [timezone]"
+---
+
+Call the MCP tool `drupal_create_user`.
+
+Create a new Drupal user account with optional roles and password.
+
+Parse the arguments supplied with this command into this tool's parameters:
+
+**Required:**
+- `name` (string): Username
+- `mail` (string): Email address
+
+**Optional:**
+- `site` (string): Named site from connector config. Omit only on reads: multi-site configs fall back to defaultSite (often local/dev, not production). Writes require an explicit site when more than one site is configured. Every response includes `_target` { name, baseUrl, source } (`hint` when you passed site, `default` when you did not).
+- `password` (string): Initial password (plaintext — sent over HTTPS)
+- `status` (boolean (true/false)): true = active (default)
+- `roles` (array (pass as JSON)): Role UUIDs to assign
+- `timezone` (string)
+
+If a required parameter is missing, ask before calling — do not invent values. Coerce each value to its JSON type (booleans → true/false, numbers → numeric, object/array → parse JSON), then make the single tool call and summarize the result.

package/.agents/commands/drupal-delete-media.md

@@ -0,0 +1,21 @@
+---
+description: "Delete a media entity. Does not delete the underlying File entity. Confirm with the user before calling."
+argument-hint: "<type> <id> [site]"
+---
+
+Call the MCP tool `drupal_delete_media`.
+
+Delete a media entity. Does not delete the underlying File entity. Confirm with the user before calling.
+
+> ⚠ **Destructive** — this permanently changes or deletes data. Confirm with the user before calling.
+
+Parse the arguments supplied with this command into this tool's parameters:
+
+**Required:**
+- `type` (string)
+- `id` (string)
+
+**Optional:**
+- `site` (string): Named site from connector config. Omit only on reads: multi-site configs fall back to defaultSite (often local/dev, not production). Writes require an explicit site when more than one site is configured. Every response includes `_target` { name, baseUrl, source } (`hint` when you passed site, `default` when you did not).
+
+If a required parameter is missing, ask before calling — do not invent values. Coerce each value to its JSON type (booleans → true/false, numbers → numeric, object/array → parse JSON), then make the single tool call and summarize the result.

package/.agents/commands/drupal-delete-node.md

@@ -0,0 +1,22 @@
+---
+description: "Permanently delete a node. Irreversible — confirm with the user before calling."
+argument-hint: "<type> <id> [site] [dryRun]"
+---
+
+Call the MCP tool `drupal_delete_node`.
+
+Permanently delete a node. Irreversible — confirm with the user before calling.
+
+> ⚠ **Destructive** — this permanently changes or deletes data. Confirm with the user before calling.
+
+Parse the arguments supplied with this command into this tool's parameters:
+
+**Required:**
+- `type` (string)
+- `id` (string): Node UUID
+
+**Optional:**
+- `site` (string): Named site from connector config. Omit only on reads: multi-site configs fall back to defaultSite (often local/dev, not production). Writes require an explicit site when more than one site is configured. Every response includes `_target` { name, baseUrl, source } (`hint` when you passed site, `default` when you did not).
+- `dryRun` (boolean (true/false)): Validate and return a preview of the delete without committing.
+
+If a required parameter is missing, ask before calling — do not invent values. Coerce each value to its JSON type (booleans → true/false, numbers → numeric, object/array → parse JSON), then make the single tool call and summarize the result.

package/.agents/commands/drupal-delete-taxonomy-term.md

@@ -0,0 +1,21 @@
+---
+description: "Delete a taxonomy term. Confirm with the user before calling."
+argument-hint: "<vocabulary> <id> [site]"
+---
+
+Call the MCP tool `drupal_delete_taxonomy_term`.
+
+Delete a taxonomy term. Confirm with the user before calling.
+
+> ⚠ **Destructive** — this permanently changes or deletes data. Confirm with the user before calling.
+
+Parse the arguments supplied with this command into this tool's parameters:
+
+**Required:**
+- `vocabulary` (string)
+- `id` (string)
+
+**Optional:**
+- `site` (string): Named site from connector config. Omit only on reads: multi-site configs fall back to defaultSite (often local/dev, not production). Writes require an explicit site when more than one site is configured. Every response includes `_target` { name, baseUrl, source } (`hint` when you passed site, `default` when you did not).
+
+If a required parameter is missing, ask before calling — do not invent values. Coerce each value to its JSON type (booleans → true/false, numbers → numeric, object/array → parse JSON), then make the single tool call and summarize the result.

package/{.claude → .agents}/commands/drupal-describe-fields.md

@@ -1,21 +1,20 @@
 ---
 description: "Introspect the fields of a Drupal entity type + bundle: returns a per-field list of { name, type, kind, cardinality?, approximate }. Read-only. Built on schema SAMPLING (an existing entity), so results are approximate — only populated fields are visible and required/cardinality/allowedValues are inferred from value shape. Authoritative field metadata comes from the Drush bridge (Field API). Use this before creating/updating entities to learn field names."
 argument-hint: "<site> [type] [entityType] [bundle]"
-allowed-tools: mcp__drupal__drupal_describe_fields
 ---
 
-Call the `mcp__drupal__drupal_describe_fields` MCP tool.
+Call the MCP tool `drupal_describe_fields`.
 
 Introspect the fields of a Drupal entity type + bundle: returns a per-field list of { name, type, kind, cardinality?, approximate }. Read-only. Built on schema SAMPLING (an existing entity), so results are approximate — only populated fields are visible and required/cardinality/allowedValues are inferred from value shape. Authoritative field metadata comes from the Drush bridge (Field API). Use this before creating/updating entities to learn field names.
 
-Parse the request in `$ARGUMENTS` into this tool's parameters:
+Parse the arguments supplied with this command into this tool's parameters:
 
 **Required:**
-- `site` (string): Configured site name.
+- `site` (string): Named site from connector config. Omit only on reads: multi-site configs fall back to defaultSite (often local/dev, not production). Writes require an explicit site when more than one site is configured. Every response includes `_target` { name, baseUrl, source } (`hint` when you passed site, `default` when you did not).
 
 **Optional:**
 - `type` (string): Entity type machine name, e.g. 'node', 'taxonomy_term', 'user', 'media'. Alias: `entityType` (as used by the sibling tools).
 - `entityType` (string): Alias for `type` — accepted for parity with get_entity_schema / entity_create / entity_update / resolve_reference.
 - `bundle` (string): Bundle machine name, e.g. 'article'. Defaults to the entity type for single-bundle types (e.g. 'user').
 
-If a required parameter is missing from `$ARGUMENTS`, ask before calling — do not invent values. Coerce each value to its JSON type (booleans → true/false, numbers → numeric, object/array → parse JSON), then make the single tool call and summarize the result.
+If a required parameter is missing, ask before calling — do not invent values. Coerce each value to its JSON type (booleans → true/false, numbers → numeric, object/array → parse JSON), then make the single tool call and summarize the result.

package/.agents/commands/drupal-drush-cache-rebuild.md

@@ -0,0 +1,15 @@
+---
+description: "Run `drush cache:rebuild` via SSH. Clears all Drupal caches. Requires drushSsh config and write access."
+argument-hint: "[site]"
+---
+
+Call the MCP tool `drupal_drush_cache_rebuild`.
+
+Run `drush cache:rebuild` via SSH. Clears all Drupal caches. Requires drushSsh config and write access.
+
+Parse the arguments supplied with this command into this tool's parameters:
+
+**Optional:**
+- `site` (string): Named site from connector config. Omit only on reads: multi-site configs fall back to defaultSite (often local/dev, not production). Writes require an explicit site when more than one site is configured. Every response includes `_target` { name, baseUrl, source } (`hint` when you passed site, `default` when you did not).
+
+If a required parameter is missing, ask before calling — do not invent values. Coerce each value to its JSON type (booleans → true/false, numbers → numeric, object/array → parse JSON), then make the single tool call and summarize the result.

package/.agents/commands/drupal-drush-config-export.md

@@ -0,0 +1,15 @@
+---
+description: "Export active configuration to the sync directory. Requires write access."
+argument-hint: "[site]"
+---
+
+Call the MCP tool `drupal_drush_config_export`.
+
+Export active configuration to the sync directory. Requires write access.
+
+Parse the arguments supplied with this command into this tool's parameters:
+
+**Optional:**
+- `site` (string): Named site from connector config. Omit only on reads: multi-site configs fall back to defaultSite (often local/dev, not production). Writes require an explicit site when more than one site is configured. Every response includes `_target` { name, baseUrl, source } (`hint` when you passed site, `default` when you did not).
+
+If a required parameter is missing, ask before calling — do not invent values. Coerce each value to its JSON type (booleans → true/false, numbers → numeric, object/array → parse JSON), then make the single tool call and summarize the result.