Your SDR agents shouldn't learn Salesforce by failing at it.
Sales agents fail in expensive, quiet ways: CRM writes that violate field-level security, duplicate contacts, sequence enrollments that double-send, deliverability burned by a misconfigured domain. Each route below is a real entry in the live corpus — the working procedure plus the gotchas that break naive attempts.
6 routes below, pulled live from the corpus · generated 2026-07-08
When your agents write to Salesforce
live corpus query: “salesforce api”
developer.salesforce.comsampled
- The API version segment (vXX.0) must match a version supported by your org — use GET /services/data/ to retrieve the list of supported versions; using a deprecated version may cause field compatibility issues.
- Case assignment rules in Salesforce are not triggered by default on API-created cases — include the 'AssignmentRuleHeader' SOAP header equivalent (via a custom header or the Apex trigger route) if you need assignment rules to fire.
salesforce.comsampled
- SOQL has no SELECT * — you must enumerate fields
- Relationship queries count child rows against the 2000-row batch differently; nextRecordsUrl is mandatory handling
When your agents manage HubSpot contacts
live corpus query: “hubspot create contact”
hubspot.comsampled
- Search index lags writes by seconds — create-after-search races produce duplicates; catch 409 with the existing id in the error body
- Custom properties must exist before you write them (create via properties API) — unknown properties 400
When your agents enroll leads in sequences
live corpus query: “outreach sequence”
dev.frontapp.comsampled
- Sending high volumes of outreach messages through Front in a short window may trigger spam filters at the receiving email provider — warm up sending volumes gradually.
- Front Sequences is a feature tied to specific plan tiers — verify your plan includes Sequences before building automation against that endpoint.
close.comsampled
- Close CRM sequence enrollments are per-lead, not per-contact; if a lead has multiple contacts, the sequence targets the lead's primary contact by default — verify the correct contact receives the emails
- A lead can only have one active subscription per sequence at a time; attempting to enroll an already-enrolled lead returns a 400 error with 'already subscribed' message rather than creating a duplicate
When deliverability decides the campaign
live corpus query: “cold email deliverability”
sendgrid.comsampled
- Sending from a non-authenticated domain falls back to via sendgrid.net and lands in spam under DMARC p=reject
- A 202 response means accepted for processing, not delivered — poll the Email Activity API or use Event Webhook
12/12 blind head-to-head wins · zero wrong routes
In the v2 benchmark (live semantic-retrieval control, 2026-07-08), the route-equipped agent won 12/12 blind comparisons with 97.2% pitfall coverage (35/36) and zero wrong-route incidents. Silent-wrong answers: 0/12 with routes vs 7/12 for the control. Full methodology, prompts, and per-task scores: waymark.network/benchmark.
Plug it into your agents
npx mcp-remote https://mcp.waymark.network/mcp
One MCP server, no key, no card — reads are free forever. Prefer raw HTTP? GET mcp.waymark.network/search?q=<task> returns the same routes as JSON.
Design-partner pilot
$750/mo · month-to-month, cancel anytime
- Priority route coverage for your agents' integration domains — you tell us where your fleet fails, we route it
- Private route namespace for your internal procedures
- 20 route-bounty credits per month included ($500 list value at the standard $25/route bounty)
- Direct line to the operator; SLA on verification
Questions first? [email protected] — you talk to the operator, not a queue. Waymark is a service of MC Software, LLC, operated by a SOC 2-audited organization.