Terminal
Terminal/Skill
Lists the technologies visible on a company's website, such as CMS, analytics and ad tools, and notes what the site cannot show.
By Clay
Works with
npx skills add clay-run/clay-skill-creator --skill detect-tech-stackThe insight: technographics answer "what can the website prove?", never "what does the company use?" — and the raw answer is a lifetime archive, not a snapshot. The managed function reads the site's source code (BuiltWith-backed), so it sees what runs in a visitor's browser and is blind to everything server-side or back-office (finance, HR, ERP always; CRM unless a form embeds it). And it returns every technology ever detected, current and historical, indistinguishably — a current Shopify Plus store still lists Magento and WordPress from years past. Both verdicts are traps: detected may be history, and not detected is never evidence of non-use.
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 |
|---|---|---|
| Domains | one hostname per row | given only a company name, resolve the domain first and sanity-check it — a wrong domain returns a confidently wrong stack with no error |
| The GTM question | map the visible stack, confirm one named tool, look below the SaaS layer, or detect a back-office competitor | ask before spending. This surface fully answers the first, answers the second only when the tool is website-visible, and only contributes partial evidence to the last two |
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 (or follow
https://raw.githubusercontent.com/clay-run/agent-plugins/main/GETTING_STARTED.md),
restart if told to, and re-run this skill. Tell the user which workspace you're in.
Company Domain (hostname). Given only a company name, resolve the domain first (a
managed Company Domain function exists) and sanity-check it: a wrong domain returns a
confidently wrong stack with no error.Use the workspace's managed Website Technology Stack function — confirm it exists
with clay routines list / get. ~2 credits per domain, and website-scan providers
charge on a miss too; state total cost and get approval before multi-domain runs
(the routines surface reports no measured billing — quote the declared rate as an
estimate). CLI envelope: {"items":[{"id":"<key>","inputs":{"Company Domain": "acme.example"}}]} via clay routines runs start ... --input -. Runs return in seconds.
For hundreds of domains or a recurring refresh, move to a table/workflow and say so.
The result is one flat comma-separated string in Website Tech Stack — no categories,
no dates. Know its three defects before reporting anything:
complete — gate on content,
never on status.Then grade each real finding:
Per domain: technologies grouped by GTM-relevant category (commerce, marketing/ads,
analytics, support, infrastructure), each as technology · evidence (corroborated / uncorroborated), a truncation note when applicable, then a not-assessable line
naming the invisible layers relevant to the user's question, then a one-line readout
tied to their stated scenario.
Ask: "Does northfield-outfitters.example run Shopify? We sell a Shopify competitor." Scenario (b), website-visible — this surface can answer. Run (~2 credits, approved): the list includes Shopify, Shopify Hosted, Shop Pay, Shopify Custom Theme → corroborated current storefront — displacement-qualified. The same list also contains "Magento": archive noise, contradicted by the corroborated Shopify family — not reported as in use. Counter-case: "Do they use Workday?" → not assessable — HR systems don't appear in website source; recommend job-postings/provider research instead of spending here.