An agent does not think it needs a data enrichment tool. It thinks it has a LinkedIn URL and needs a job title. This view is the closed vocabulary of 55 jobs in 10 families, each one a verb plus an object, phrased from the agent's side. 827 tags are assigned across 271 of 293 entries. 22 entries carry no tag at all and each one records why.
Turn a criterion, a domain, a URL or an email into a record. 7 jobs, 97 tagged entries.
Turn a person into a way to reach them, and check it is real. 3 jobs, 63 tagged entries.
Find out something changed, or find out what is true right now. 9 jobs, 117 tagged entries.
Write it, send it, dial it, and read back what happened. 7 jobs, 184 tagged entries.
Get a meeting on the calendar and get the words out of it. 7 jobs, 127 tagged entries.
Read, write and move the data the rest of the stack argues about. 6 jobs, 79 tagged entries.
Quote it, paper it, sign it, and share it with the buyer. 6 jobs, 31 tagged entries.
Decide what the number is and who is going to hit it. 5 jobs, 89 tagged entries.
Own the mailboxes, the reputation and the placement. 3 jobs, 26 tagged entries.
Find servers, or stand between an agent and somebody's SaaS. 2 jobs, 14 tagged entries.
Vocabulary source: data/jobs.yaml, sha256 e63c27779ba7bdea... Tags source: data/tags.yaml, sha256 77a593d1de719f44... The vocabulary is closed: a tool whose job is not in it stays blank rather than being forced into the nearest tag.