메뉴
Generate use-case-specific Attio workflow skills from templates. Use when creating new skills for lead qualification, deal management, customer onboarding, or custom Attio workflows.
기본적으로 소스를 먼저 확인하는 Prompt가 선택됩니다. 직접 명령으로 전환하거나 로컬 사본을 다운로드할 수도 있습니다.
설치 여부를 결정하기 전에 SKILL.md와 SkillsMP에 표시된 보조 파일을 읽어 보세요.
Codex 또는 Claude로 설치 이 Prompt를 복사해 Codex, Claude 또는 다른 어시스턴트에 붙여 넣으면 Skill 페이지를 검토하고 설치를 진행할 수 있습니다.
직접 명령은 검토 Prompt를 거치지 않습니다. 실행하기 전에 소스를 확인하세요.
npx skills add https://github.com/kesslerio/attio-mcp-server --skill attio-skill-generator명령은 한 줄로 유지됩니다. 복사하기 전에 가로로 스크롤해 전체 내용을 확인하세요.
로컬 사본을 원하시나요? SkillsMP에서 현재 제공할 수 있는 파일을 다운로드하세요.
SOC 직업 분류 기준
SKILL.md 표시 중
| name | attio-skill-generator |
| description | Generate use-case-specific Attio workflow skills from templates. Use when creating new skills for lead qualification, deal management, customer onboarding, or custom Attio workflows. |
| license | Apache-2.0 (see LICENSE.txt for attribution) |
A meta-skill that generates customized Attio workflow skills tailored to your workspace.
Use this skill when you want to:
| Use Case | Primary Object | Related Objects | Description |
|---|---|---|---|
lead-qualification | companies | people | Qualify and score inbound leads |
deal-management | deals | companies, people | Manage deals through pipeline stages |
customer-onboarding | companies | people, deals | Structured onboarding workflows |
You (Claude) must discover the workspace schema for the specific use-case. The Python scripts are sandboxed and cannot access APIs or MCP tools.
IMPORTANT: Gather data for the PRIMARY OBJECT only. Do not gather full schemas for all objects.
Determine what to gather based on use-case:
| Use Case | Gather Attributes For | Gather Lists? |
|---|---|---|
lead-qualification | companies | NO (unless requested) |
deal-management | deals | NO (unless requested) |
customer-onboarding | companies | NO (unless requested) |
FIRST: Check for attio-workspace-schema skill
Before using MCP tools, check if the attio-workspace-schema skill is installed by looking for its resource files.
Option A: Use attio-workspace-schema skill (PREFERRED - faster, no API calls)
If attio-workspace-schema skill exists:
deal-management → read resources/deals-attributes.mdlead-qualification → read resources/companies-attributes.mdcustomer-onboarding → read resources/companies-attributes.mdOption B: Query MCP tools (FALLBACK - only if no schema skill)
If attio-workspace-schema skill is NOT available:
1. Call records_discover_attributes for the PRIMARY OBJECT ONLY
- deal-management → records_discover_attributes for "deals"
- lead-qualification → records_discover_attributes for "companies"
- customer-onboarding → records_discover_attributes for "companies"
2. For select/status fields on the primary object, call records_get_attribute_options
Do NOT call get-lists unless the user specifically asks for list-related functionality. Lists are organizational containers, not essential to core workflows like deal management or lead qualification.
Do NOT gather:
Note on related objects: Deal management involves linked companies and people, but these relationships are handled through record-reference fields on the deals object. You don't need full attribute lists for related objects.
Structure the discovered data as JSON for the generator. This structure is CRITICAL for correct output.
⚠️ REQUIRED FIELDS for each attribute:
api_slug - The API field name (required)type - Field type: text, status, select, number, date, etc. (required)is_required - Boolean, true if field is required (optional, defaults to false)is_multiselect - Boolean, true if field accepts multiple values (optional, defaults to false)options - Array of option objects for status/select fields (REQUIRED for status/select types!)⚠️ OPTIONS ARE CRITICAL: For status and select type fields, you MUST include the options array with the actual option titles from the workspace. Without this, the generated skill won't show pipeline stages or dropdown values!
{
"objects": {
"deals": {
"display_name": "Deals",
"attributes": [
{
"api_slug": "name",
"display_name": "Deal Name",
"type": "text",
"is_required": true,
"is_multiselect": false
},
{
"api_slug": "stage",
"display_name": "Deal Stage",
"type": "status",
"is_required": true,
"is_multiselect": false,
"options": [
{ "title": "MQL" },
{ "title": "Demo Request" },
{ "title": "Discovery Call" },
{ "title": "Demo Booked" },
{ "title": "Negotiations" },
{ "title": "Won 🎉" },
{ "title": "Lost" }
]
},
{
"api_slug": "primary_interest",
"display_name": "Primary Interest",
"type": "select",
"is_multiselect": false,
"options": [
{ "title": "GLP-1 / medical weight loss" },
{ "title": "Body contouring / aesthetics" },
{ "title": "Replacing existing device" }
]
},
{
"api_slug": "lost_reason",
"display_name": "Lost Reason",
"type": "select",
"is_multiselect": true,
"options": [
{ "title": "Pricing/Cost" },
{ "title": "Competitor" },
{ "title": "Timing Not Right" }
]
},
{
"api_slug": "associated_people",
"display_name": "Associated People",
"type": "record-reference",
"is_multiselect": true
},
{
"api_slug": "associated_company",
"display_name": "Company",
"type": "record-reference",
"is_multiselect": false
}
]
}
},
"lists": []
}
Key points:
lists array is empty by default. Only populate if user requests list functionality.status/select fields, copy the EXACT option titles from the workspace schemais_multiselect: true for fields that accept multiple values (check the Multi column in schema)Execute the generator script with the workspace schema.
Recommended: Use file-based input (avoids shell escaping issues with large JSON):
# Save schema to file first
echo '<JSON from Step 2>' > workspace-schema.json
# Run generator with file input
python scripts/generator.py \
--use-case lead-qualification \
--name acme-lead-qualification \
--workspace-schema-file workspace-schema.json \
--output ./generated-skills
Alternative: Inline JSON (only for small schemas):
python scripts/generator.py \
--use-case lead-qualification \
--name acme-lead-qualification \
--workspace-schema '{"objects": {...}}' \
--output ./generated-skills
Parameters:
--use-case: One of lead-qualification, deal-management, customer-onboarding--name: Skill name (hyphen-case, max 64 chars)--workspace-schema-file: Path to JSON file with workspace data (recommended)--workspace-schema: JSON string with workspace data (alternative for small schemas)--output: Output directory (default: ./generated-skills)Show the user the generated SKILL.md content for review:
cat ./generated-skills/acme-lead-qualification/SKILL.md
Allow the user to request modifications before packaging.
Validate the generated skill:
python scripts/quick_validate.py ./generated-skills/acme-lead-qualification
Package as a .skill file:
python scripts/package_skill.py ./generated-skills/acme-lead-qualification
Provide the user with:
.skill ZIP file| Script | Purpose | Input |
|---|---|---|
generator.py | Generate skill from templates | JSON schema + use-case |
init_skill.py | Initialize empty skill structure | Skill name |
package_skill.py | Validate and create ZIP | Skill directory path |
quick_validate.py | Validate SKILL.md frontmatter | Skill directory path |
User: "Use attio-skill-generator to create a Deal Management skill for my workspace"
Claude:
resources/deals-attributes.md]records_discover_attributes for deals only]get-lists - lists are not needed for deal management]python scripts/generator.py --use-case deal-management --name my-deal-management --workspace-schema-file schema.json
python scripts/package_skill.py ./generated-skills/my-deal-management
./my-deal-management.skillNote: The skill includes record-reference fields (associated_company, contacts) that link to companies and people, but without documenting those objects' full schemas. Lists are not included unless explicitly requested.
Templates are in resources/templates/. You can customize:
SKILL.template.md - Main skill metadata and structureworkflows.template.md - Workflow step patternstool-reference.template.md - MCP tool referenceexamples.template.md - Example interactionsUse-case configurations are in resources/use-cases/:
lead-qualification.yamldeal-management.yamlcustomer-onboarding.yamlThe generator requires these Python packages:
chevron - Handlebars-compatible templatingpyyaml - YAML parsingInstall if needed: pip install chevron pyyaml
Before packaging, verify the generated skill:
See resources/validation-checklist.md for detailed checklist.
This skill extends anthropics/skills/skill-creator. Licensed under Apache License 2.0. See LICENSE.txt for full terms.
Modifications from original: