Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Automate SendGrid email operations including sending emails, managing contacts/lists, sender identities, templates, and analytics via Rube MCP (Composio). Always search tools first for current schemas.
.claude/skills/davepoon-sendgrid-automation/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-10 | ✗→✓ | ▲ Improved | 338% | 0% |
| case-09 | ✗→✓ | ▲ Improved | 282% | 0% |
| case-13 | ✗→✓ | ▲ Improved | 103% | 0% |
| case-14 | ✗→✓ | ▲ Improved | 102% | 0% |
| case-15 | ✗→✓ | ▲ Improved | 241% | 0% |
Automate SendGrid email delivery workflows including marketing campaigns (Single Sends), contact and list management, sender identity setup, and email analytics through Composio's SendGrid toolkit.
Toolkit docs: composio.dev/toolkits/sendgrid
RUBE_MANAGE_CONNECTIONS with toolkit sendgridRUBE_SEARCH_TOOLS first to get current tool schemasGet Rube MCP: Add https://rube.app/mcp as an MCP server in your client configuration. No API keys needed — just add the endpoint and it works.
RUBE_SEARCH_TOOLS respondsRUBE_MANAGE_CONNECTIONS with toolkit sendgridWhen to use: User wants to create and send a marketing email campaign to a contact list or segment.
Tool sequence:
SENDGRID_RETRIEVE_ALL_LISTS - List available marketing lists to target Prerequisite]SENDGRID_CREATE_A_LIST - Create a new list if needed Optional]SENDGRID_ADD_OR_UPDATE_A_CONTACT - Add contacts to the list Optional]SENDGRID_GET_ALL_SENDER_IDENTITIES - Get verified sender ID Prerequisite]SENDGRID_CREATE_SINGLE_SEND - Create the campaign with content, sender, and recipients Required]Key parameters for SENDGRID_CREATE_SINGLE_SEND:
name: Campaign name (required)email__config__subject: Email subject lineemail__config__html__content: HTML body contentemail__config__plain__content: Plain text versionemail__config__sender__id: Verified sender identity IDemail__config__design__id: Use instead of html_content for pre-built designssend__to__list__ids: Array of list UUIDs to send tosend__to__segment__ids: Array of segment UUIDssend__to__all: true to send to all contactsemail__config__suppression__group__id or email__config__custom__unsubscribe__url: One required for compliancePitfalls:
send_at on CREATE does NOT schedule the send; it only prepopulates the UI date; use the Schedule endpoint separatelysend_at: "now" is only valid with the Schedule endpoint, not CREATEsuppression_group_id or custom_unsubscribe_url for unsubscribe complianceSENDGRID_GET_ALL_SENDER_IDENTITIESemail__config__subject)When to use: User wants to create contact lists, add/update contacts, search for contacts, or remove contacts from lists.
Tool sequence:
SENDGRID_RETRIEVE_ALL_LISTS - List all marketing lists Required]SENDGRID_CREATE_A_LIST - Create a new contact list Optional]SENDGRID_GET_A_LIST_BY_ID - Get list details and sample contacts Optional]SENDGRID_ADD_OR_UPDATE_A_CONTACT - Upsert contacts with list association Required]SENDGRID_GET_CONTACTS_BY_EMAILS - Look up contacts by email Optional]SENDGRID_GET_CONTACTS_BY_IDENTIFIERS - Look up contacts by email, phone, or external ID Optional]SENDGRID_GET_LIST_CONTACT_COUNT - Verify contact count after operations Optional]SENDGRID_REMOVE_CONTACTS_FROM_A_LIST - Remove contacts from a list without deleting Optional]SENDGRID_REMOVE_LIST_AND_OPTIONAL_CONTACTS - Delete an entire list Optional]SENDGRID_IMPORT_CONTACTS - Bulk import from CSV Optional]Key parameters for SENDGRID_ADD_OR_UPDATE_A_CONTACT:
contacts: Array of contact objects (max 30,000 or 6MB), each with at least one identifier: email, phone_number_id, external_id, or anonymous_id (required)list_ids: Array of list UUID strings to associate contacts withPitfalls:
SENDGRID_ADD_OR_UPDATE_A_CONTACT is asynchronous; returns 202 with job_id; contacts may take 10-30 seconds to appearSENDGRID_ADD_A_SINGLE_RECIPIENT_TO_A_LIST uses the legacy API; prefer SENDGRID_ADD_OR_UPDATE_A_CONTACT with list_idsSENDGRID_REMOVE_LIST_AND_OPTIONAL_CONTACTS is irreversible; require explicit user confirmationWhen to use: User wants to set up or view sender identities (From addresses) for sending emails.
Tool sequence:
SENDGRID_GET_ALL_SENDER_IDENTITIES - List all existing sender identities Required]SENDGRID_CREATE_A_SENDER_IDENTITY - Create a new sender identity Optional]SENDGRID_VIEW_A_SENDER_IDENTITY - View details for a specific sender Optional]SENDGRID_UPDATE_A_SENDER_IDENTITY - Update sender details Optional]SENDGRID_CREATE_VERIFIED_SENDER_REQUEST - Create and verify a new sender Optional]SENDGRID_AUTHENTICATE_A_DOMAIN - Set up domain authentication for auto-verification Optional]Key parameters for SENDGRID_CREATE_A_SENDER_IDENTITY:
from__email: From email address (required)from__name: Display name (required)reply__to__email: Reply-to address (required)nickname: Internal identifier (required)address, city, country: Physical address for CAN-SPAM compliance (required)Pitfalls:
SENDGRID_CREATE_VERIFIED_SENDER_REQUEST sends a verification email; sender is unusable until verifiedWhen to use: User wants to review email delivery stats, bounce rates, open/click metrics, or message activity.
Tool sequence:
SENDGRID_RETRIEVE_GLOBAL_EMAIL_STATISTICS - Get account-wide delivery metrics Required]SENDGRID_GET_ALL_CATEGORIES - Discover available categories for filtering Optional]SENDGRID_RETRIEVE_EMAIL_STATISTICS_FOR_CATEGORIES - Get stats broken down by category Optional]SENDGRID_FILTER_ALL_MESSAGES - Search email activity feed by recipient, status, or date Optional]SENDGRID_FILTER_MESSAGES_BY_MESSAGE_ID - Get detailed events for a specific message Optional]SENDGRID_REQUEST_CSV - Export activity data as CSV for large datasets Optional]SENDGRID_DOWNLOAD_CSV - Download the exported CSV file Optional]Key parameters for SENDGRID_RETRIEVE_GLOBAL_EMAIL_STATISTICS:
start_date: Start date YYYY-MM-DD (required)end_date: End date YYYY-MM-DDaggregated_by: "day", "week", or "month"limit / offset: Pagination (default 500)Key parameters for SENDGRID_FILTER_ALL_MESSAGES:
query: SQL-like query string, e.g., status="delivered", to_email="user@example.com", date ranges with BETWEEN TIMESTAMPlimit: 1-1000 (default 10)Pitfalls:
SENDGRID_FILTER_ALL_MESSAGES requires the "30 Days Additional Email Activity History" paid add-on; returns 403 without itdetails[].stats[0].metrics, not a flat structureSENDGRID_RETRIEVE_EMAIL_STATISTICS_FOR_CATEGORIESWhen to use: User wants to check or manage unsubscribe groups for email compliance.
Tool sequence:
SENDGRID_GET_SUPPRESSION_GROUPS - List all suppression groups Required]SENDGRID_RETRIEVE_ALL_SUPPRESSION_GROUPS_FOR_AN_EMAIL_ADDRESS - Check suppression status for a specific email Optional]Pitfalls:
Always resolve names to IDs before operations:
SENDGRID_RETRIEVE_ALL_LISTS and match by nameSENDGRID_GET_ALL_SENDER_IDENTITIES and matchSENDGRID_GET_CONTACTS_BY_EMAILS with email arraySENDGRID_RETRIEVE_ALL_LISTS: Token-based with page_token and page_size (max 1000)SENDGRID_RETRIEVE_GLOBAL_EMAIL_STATISTICS: Offset-based with limit (max 500) and offsetContact operations (ADD_OR_UPDATE_A_CONTACT, IMPORT_CONTACTS) are asynchronous:
job_idGET_CONTACTS_BY_EMAILSGET_LIST_CONTACT_COUNT to confirm list growthRetry-After headersemail__config__subject, from__emailsend_at on CREATE_SINGLE_SEND only sets a UI default, does NOT scheduleSENDGRID_ADD_A_SINGLE_RECIPIENT_TO_A_LIST uses legacy API; recipient_id is Base64-encoded lowercase emailSENDGRID_RETRIEVE_ALL_LISTS and SENDGRID_GET_ALL_LISTS both exist; prefer RETRIEVE_ALL_LISTS for Marketing API/v3/contactdb/) which may return 403 on newer accountsSENDGRID_ADD_OR_UPDATE_A_CONTACT, SENDGRID_RETRIEVE_ALL_LISTS, SENDGRID_CREATE_SINGLE_SEND| Task | Tool Slug | Key Params | |------|-----------|------------| | List marketing lists | SENDGRID_RETRIEVE_ALL_LISTS | page_size, page_token | | Create list | SENDGRID_CREATE_A_LIST | name | | Get list by ID | SENDGRID_GET_A_LIST_BY_ID | id | | Get list count | SENDGRID_GET_LIST_CONTACT_COUNT | id | | Add/update contacts | SENDGRID_ADD_OR_UPDATE_A_CONTACT | contacts, list_ids | | Search contacts by email | SENDGRID_GET_CONTACTS_BY_EMAILS | emails | | Search by identifiers | SENDGRID_GET_CONTACTS_BY_IDENTIFIERS | identifier_type, identifiers | | Remove from list | SENDGRID_REMOVE_CONTACTS_FROM_A_LIST | id, contact_ids | | Delete list | SENDGRID_REMOVE_LIST_AND_OPTIONAL_CONTACTS | id, delete_contacts | | Import contacts CSV | SENDGRID_IMPORT_CONTACTS | field mappings | | Create Single Send | SENDGRID_CREATE_SINGLE_SEND | name, email__config__*, send__to__list__ids | | List sender identities | SENDGRID_GET_ALL_SENDER_IDENTITIES | (none) | | Create sender | SENDGRID_CREATE_A_SENDER_IDENTITY | from__email, from__name, address | | Verify sender | SENDGRID_CREATE_VERIFIED_SENDER_REQUEST | from_email, nickname, address | | Authenticate domain | SENDGRID_AUTHENTICATE_A_DOMAIN | domain | | Global email stats | SENDGRID_RETRIEVE_GLOBAL_EMAIL_STATISTICS | start_date, aggregated_by | | Category stats | SENDGRID_RETRIEVE_EMAIL_STATISTICS_FOR_CATEGORIES | start_date, categories | | Filter email activity | SENDGRID_FILTER_ALL_MESSAGES | query, limit | | Message details | SENDGRID_FILTER_MESSAGES_BY_MESSAGE_ID | msg_id | | Export CSV | SENDGRID_REQUEST_CSV | query | | Download CSV | SENDGRID_DOWNLOAD_CSV | download_uuid | | List categories | SENDGRID_GET_ALL_CATEGORIES | (none) | | Suppression groups | SENDGRID_GET_SUPPRESSION_GROUPS | (none) | | Get template | SENDGRID_RETRIEVE_A_SINGLE_TRANSACTIONAL_TEMPLATE | template_id | | Duplicate template | SENDGRID_DUPLICATE_A_TRANSACTIONAL_TEMPLATE | template_id, name |
Powered by Composio
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-10 | fail→pass | 6,773 | 4,614 | -32% | 1 | 1 | 0% | 1,057 | 4,632 | +338% | 0 | 0 | — |
case-09 | fail→pass | 6,481 | 4,754 | -27% | 1 | 1 | 0% | 1,194 | 4,559 | +282% | 0 | 0 | — |
case-01 | pass→fail | 12,877 | 3,946 | -69% | 1 | 1 | 0% | 2,641 | 3,923 | +49% | 0 | 0 | — |
case-02 | pass→pass | 11,243 | 8,369 | -26% | 1 | 1 | 0% | 2,339 | 5,350 | +129% | 0 | 0 | — |
case-03 | pass→fail | 8,991 | 5,072 | -44% | 1 | 1 | 0% | 1,689 | 3,954 | +134% | 0 | 0 | — |
case-04 | fail→fail | 9,097 | 17,380 | +91% | 1 | 1 | 0% | 1,856 | 4,147 | +123% | 0 | 0 | — |
case-05 | fail→fail | 7,335 | 6,099 | -17% | 1 | 1 | 0% | 626 | 4,161 | +565% | 0 | 0 | — |
case-06 | fail→fail | 8,961 | 6,720 | -25% | 1 | 1 | 0% | 621 | 4,224 | +580% | 0 | 0 | — |
case-07 | pass→pass | 9,102 | 4,856 | -47% | 1 | 1 | 0% | 1,707 | 4,539 | +166% | 0 | 0 | — |
case-08 | pass→pass | 9,261 | 5,502 | -41% | 1 | 1 | 0% | 1,719 | 4,558 | +165% | 0 | 0 | — |
case-11 | pass→pass | 8,699 | 4,464 | -49% | 1 | 1 | 0% | 1,531 | 4,497 | +194% | 0 | 0 | — |
case-12 | pass→pass | 7,398 | 3,321 | -55% | 1 | 1 | 0% | 1,454 | 4,345 | +199% | 0 | 0 | — |
case-13 | fail→pass | 12,217 | 3,905 | -68% | 1 | 1 | 0% | 2,169 | 4,393 | +103% | 0 | 0 | — |
case-14 | fail→pass | 13,443 | 5,166 | -62% | 1 | 1 | 0% | 2,299 | 4,640 | +102% | 0 | 0 | — |
case-15 | fail→pass | 6,825 | 4,603 | -33% | 1 | 1 | 0% | 1,359 | 4,628 | +241% | 0 | 0 | — |
case-16 | fail→fail | 3,743 | 6,343 | +69% | 1 | 1 | 0% | 728 | 4,164 | +472% | 0 | 0 | — |
case-17 | fail→pass | 11,966 | 3,245 | -73% | 1 | 1 | 0% | 2,192 | 4,369 | +99% | 0 | 0 | — |
case-18 | fail→fail | 7,024 | 5,707 | -19% | 1 | 1 | 0% | 1,556 | 3,960 | +154% | 0 | 0 | — |
case-19 | fail→pass | 6,012 | 2,604 | -57% | 1 | 1 | 0% | 831 | 4,120 | +396% | 0 | 0 | — |
case-20 | pass→pass | 10,699 | 3,653 | -66% | 1 | 1 | 0% | 1,788 | 4,381 | +145% | 0 | 0 | — |
case-21 | fail→pass | 11,735 | 6,871 | -41% | 1 | 1 | 0% | 1,935 | 4,893 | +153% | 0 | 0 | — |
case-22 | pass→fail | 9,813 | 6,320 | -36% | 1 | 1 | 0% | 2,046 | 4,100 | +100% | 0 | 0 | — |
DecimalAI ran this skill against gemini-3.6-flash twice over the same eval suite — once with the skill loaded and once without — and compared the two runs case by case. 22 cases were attempted, and 14 counted toward the lift figure. The other 8 produced results that are not comparable between the two arms, so they are excluded from the headline rather than averaged into it. The headline lift of +23 percentage points is the difference between those two pass rates over the 14 comparable cases. 3 cases got worse with the skill loaded, and they are included in that figure.
Without the skill loaded, the model failed this case. With it loaded, the same prompt on the same model passed. This is one improved case from the latest verified run; every case, including any that regressed, is in the table above.
Other measured skills in the registry, with their headline benchmark lift.