MedBridge Sales Ops
Core Workflow
- Read the user prompt and the provided
input/payloads/answer_template.json. Treat the template as the output contract: preserve keys, arrays, enum spellings, date format, nullability, and numeric precision style.
- Identify the task family:
- Quote or RFQ pricing: prompt mentions a quote/RFQ, products/modules, EXW pricing, freight, payment terms, or quote policies.
- Engagement reconciliation: prompt mentions an opportunity, milestones/phases, invoices, payments, revenue journals, events, vouchers, or follow-up actions.
- Resolve the task API base URL from the prompt or runner-provided environment. Use only the MedBridge API records needed for the IDs in the prompt/template.
- Gather linked records with direct endpoints and ID-specific search. Prefer exact IDs from the prompt. Avoid broad listing unless direct fetch/search cannot identify a required linked record.
- Apply the derivation rules in references/derivation-rules.md.
- Return only valid JSON. Do not include markdown, comments, provenance notes, or extra keys.
API Gathering
The API root is /api; useful collections include customers, products, rfqs, quotes, freight-quotes, policies, opportunities, invoices, payments, revenue-journals, events, and vouchers.
Use direct GETs when you know the ID:
curl -sS "$BASE_URL/api/quotes/$QUOTE_ID"
curl -sS "$BASE_URL/api/products/$PRODUCT_CODE"
curl -sS "$BASE_URL/api/customers/$CUSTOMER_ID"
Use search to collect linked records when the API does not expose a filtered list:
curl -sS "$BASE_URL/api/search?q=$QUOTE_ID"
curl -sS "$BASE_URL/api/search?q=$OPPORTUNITY_ID"
curl -sS "$BASE_URL/api/search?q=$EVENT_ID"
Optional helper: run scripts/collect_context.py to extract IDs from a prompt and perform ID-specific searches:
python skill/scripts/collect_context.py --base-url "$BASE_URL" --prompt input/prompt.txt
Output Discipline
- Build the response by filling the template, not by inventing a new schema.
- Use API values over prompt wording when they conflict, except for explicit user-confirmed quantities/dates in the prompt.
- Keep line items at the requested granularity. For module RFQs, quote module product codes and ignore component composition notes unless the template explicitly asks for components.
- Use uppercase enum values exactly as declared by the template.
- Use
null where the template permits null and the value is not applicable.
1---2name: medbridge-sales-ops-123description: Solve MedBridge Sales Ops API tasks that require account-ready JSON for quote pricing, freight comparisons, RFQs, opportunity milestone reconciliation, revenue recognition, invoice follow-up, events, and vouchers. Use when prompts mention MedBridge Sales Ops, quote/RFQ/opportunity IDs, product catalog tiers, freight quotes, policies, or returning JSON matching an answer_template.json.4metadata:5 source: https://skillmd.com/skills/prism-shadow/medbridge-sales-ops-126---78# MedBridge Sales Ops910## Core Workflow11121. Read the user prompt and the provided `input/payloads/answer_template.json`. Treat the template as the output contract: preserve keys, arrays, enum spellings, date format, nullability, and numeric precision style.132. Identify the task family:14 - Quote or RFQ pricing: prompt mentions a quote/RFQ, products/modules, EXW pricing, freight, payment terms, or quote policies.15 - Engagement reconciliation: prompt mentions an opportunity, milestones/phases, invoices, payments, revenue journals, events, vouchers, or follow-up actions.163. Resolve the task API base URL from the prompt or runner-provided environment. Use only the MedBridge API records needed for the IDs in the prompt/template.174. Gather linked records with direct endpoints and ID-specific search. Prefer exact IDs from the prompt. Avoid broad listing unless direct fetch/search cannot identify a required linked record.185. Apply the derivation rules in [references/derivation-rules.md](references/derivation-rules.md).196. Return only valid JSON. Do not include markdown, comments, provenance notes, or extra keys.2021## API Gathering2223The API root is `/api`; useful collections include `customers`, `products`, `rfqs`, `quotes`, `freight-quotes`, `policies`, `opportunities`, `invoices`, `payments`, `revenue-journals`, `events`, and `vouchers`.2425Use direct GETs when you know the ID:2627```bash28curl -sS "$BASE_URL/api/quotes/$QUOTE_ID"29curl -sS "$BASE_URL/api/products/$PRODUCT_CODE"30curl -sS "$BASE_URL/api/customers/$CUSTOMER_ID"31```3233Use search to collect linked records when the API does not expose a filtered list:3435```bash36curl -sS "$BASE_URL/api/search?q=$QUOTE_ID"37curl -sS "$BASE_URL/api/search?q=$OPPORTUNITY_ID"38curl -sS "$BASE_URL/api/search?q=$EVENT_ID"39```4041Optional helper: run [scripts/collect_context.py](scripts/collect_context.py) to extract IDs from a prompt and perform ID-specific searches:4243```bash44python skill/scripts/collect_context.py --base-url "$BASE_URL" --prompt input/prompt.txt45```4647## Output Discipline4849- Build the response by filling the template, not by inventing a new schema.50- Use API values over prompt wording when they conflict, except for explicit user-confirmed quantities/dates in the prompt.51- Keep line items at the requested granularity. For module RFQs, quote module product codes and ignore component composition notes unless the template explicitly asks for components.52- Use uppercase enum values exactly as declared by the template.53- Use `null` where the template permits null and the value is not applicable.