Terminal
Terminal/Skill
Finds a person's LinkedIn profile URL from their name and company or email, and checks it against the employer before reporting it.
By Clay
Works with
npx skills add clay-run/clay-skill-creator --skill find-linkedin-profileThe insight: a returned URL is a candidate, not an answer — and a rejected candidate is not a dead end. Finders hand back plausible URLs for the wrong person without erroring; slugs go stale and vary; a reject usually means a same-name collision, not a miss. The ladder: find → validate → recover — a wrong profile is worse than none.
Nothing here ships with a value. Each one is the installer's, not the author's: ask for it, never substitute a plausible default, and where an answer does not exist say which step becomes unavailable rather than guessing. Where a default IS defensible it is named below, and using it means saying so in the output.
| Input | What the installer supplies | If it is missing |
|---|---|---|
| What they have per row | a profile URL, an email, or a name plus company | no default — the route depends on it, and each route has a different failure mode |
| The expected employer | for personal-email rows, who they think the person works for | ask. A personal email carries no company anchor, so without this the match is unverifiable |
| Budget | credits, on the rows that need enrichment | rows that already carry a URL cost nothing; state the count that does |
If an answer sheet is present beside this skill, load it and ask only for what it does not cover. A partial sheet is normal; a value it is missing gets asked for on its own rather than restarting the interview. Say which values came from the sheet before using them — a sheet applied silently is a wrong field nobody catches. If there is no sheet, say nothing about sheets — the check is a file lookup, not a question, so run the interview as though the feature did not exist rather than reporting an absence. At delivery, offer to save the answers back (identifiers only — never a token or a password), private and never published — and phrase the offer so it explains itself: "want me to save your answers to a file, so the next person on your team doesn't have to answer these again?"
Run clay whoami; echo "exit_code=$?". If it fails or Clay tools are missing, run the Clay
plugin's setup skill and re-run this skill. Tell the user which workspace you're in.
clay search filters-mode create --source-type people \
--filters '{"names":["<full name>"],"company_identifier":["<domain>"]}'
clay search filters-mode run <searchId>
company_identifier must be an array. Records return url, name, current title, and role
start date; this spends search-result quota, not credits. Gate on match count:
"include_past_experiences": true. Still 0 → report honestly.job_title_keywords or location if given;
otherwise show the name/title list and ask (hasMore: true = still more).Enrich the candidate with the managed Enrich Person function (clay routines list /
get to confirm — the list paginates). It accepts Professional Profile URL or Email
only — NOT name+company, despite its description. Envelope:
{"items":[{"id":"<key>","inputs":{"Professional Profile URL":"<candidate>"}}]} via
clay routines runs start. Score three signals:
/in/priya-r for a surname starting with R).All three → validated: report the payload's url — the canonical slug, never the raw
search hit (sources disagree on slugs for one person). Two → low confidence; name the
failed signal. Name fails or fewer → rejected → Step 4. Empty result: {} with
status: complete = not found — completion is not data (not-founds return in seconds
here — no waterfall; a slow run signals trouble, not a miss). Name matches,
employer doesn't → moved: right person, new company — report URL + current employer +
flag. A stored URL that re-enriches empty = "could not re-verify" — recover; never
overwrite the stored value with nothing.
A dead slug or a rejected candidate → re-run Step 2 with a disambiguator (title, location). The recovered URL MUST differ from the rejected one and MUST pass Step 3 itself. No decisively better candidate → not found / ambiguous — empty beats wrong.
The deliverable: the canonical enriched URL plus confidence and 1–3 short reasons quoting the validating facts. Ambiguity is surfaced, never resolved by rank. The common mistake: trusting one confident-looking hit — a single match can still be a same-named stranger.
Per person: name · company · LinkedIn URL · status (validated / low confidence / ambiguous — n candidates / moved — now at X / not found) · reasons (1–3 short strings).
Batches get a summary line: validated %, low-confidence %, ambiguous %, not found %.
Ask: "Find the LinkedIn for Dana Whitfield at brightloop.example." One hit,
/in/dana-whitfield-8a41b2 · "VP Operations". Enrich → canonical /in/danawhitfield, org
Brightloop, full history → validated ("VP Operations at Brightloop; name exact").
Counter-example: "Alex Rivera at meridianbank.example" → 6 profiles, different titles →
ambiguous — 6 candidates with the title list — never the first one.