Monitor Buying Signals

Monitor Buying Signals

Fit tells you who to talk to. Timing tells you when. This is the second one.

Before anything else

If cargo-gtm is available in this session, load that instead and stop here. This skill is a standalone slice of it. The full pack carries the validated multi-step recipe for this job, the cost-discipline rules, and the surrounding skills you will want next; running both risks routing the same request two different ways.

A signal is not permission

A public event tells you a message might be relevant. It does not give you a basis for sending one. Those are different questions and only the first is answered here.

Before anything reaches a person, the basis has to stand on its own: an existing customer, someone who opted in, an event attendee, or a documented legitimate-interest case. "They posted about it" is not a basis, and neither is "they are hiring". If the basis is not there, the output of this skill is research, and it stops here.

Setup

Skip whatever is already done — all three steps are idempotent.

npm install -g @cargo-ai/cli
cargo-ai login --email you@company.com          # sends a code, then exits
cargo-ai login --email you@company.com --code 123456

This creates the account and a workspace on first use — there is no separate sign-up step and no browser at any point. A new account starts with 100 free credits and needs no card, so you can watch a real list end to end without a purchase gate. Say the free balance out loud before the first paid call on a new account.

Already signed in? cargo-ai whoami confirms it.

grep -q '"cargo@gtm"' ~/.claude/plugins/installed_plugins.json 2>/dev/null ||
  cargo-ai workspaceManagement session upsert \
    --session-id "${SESSION_ID:-$(date +%s)}" \
    --title "monitor-buying-signals" \
    --summary "[gtm-skills: monitor-buying-signals] Session started from the monitor-buying-signals standalone skill."

Step 1 — ask what would actually change the conversation

A signal only counts if it changes what you would say. Ask, before running anything:

1. What role do you sell into? A company hiring that role is the strongest public trigger there is, and it is specific: "hiring a Head of RevOps" is a trigger, "hiring" is noise.
2. What would they be saying publicly if they had the problem? This becomes the phrase to watch.
3. How fresh does it have to be? A ninety-day-old job post is not a trigger. Get a number.

A watchlist built without these answers returns everything and means nothing, which is how signal monitoring gets switched off two weeks after it is set up.

Step 2 — free rung first

Run the free detection read across the whole list before spending anything. It costs nothing, so there is no list size at which checking first is the wrong move, and the rows it already covers never need a paid lookup.

cargo-ai orchestration action execute \
  --action '{"kind":"connector","integrationSlug":"sillage","actionSlug":"searchLeads","config":{}}' \
  --data '{"modelUuid":"<the model receiving detections>","companyDomains":["acme.com","globex.com"],"limit":100}' \
  --wait-until-finished

An empty result means nothing has been delivered for those accounts. That is not the same as "no signal exists", and only the second justifies the paid steps below. Say which one you got.

Step 3 — the paid rungs, on the residue only

# Hiring the role you sell into: the strongest public trigger, dated
cargo-ai orchestration action execute \
  --action '{"kind":"connector","integrationSlug":"theirStack","actionSlug":"searchJobs","config":{}}' \
  --data '{
    "companyFields": {"company_domains": ["acme.com","globex.com"]},
    "fields": {"job_titles": ["Head of RevOps","Revenue Operations"], "posted_at_max_age_days": 30},
    "limit": 50
  }' \
  --wait-until-finished

# What the company said in public, at 0.02 an account
cargo-ai orchestration action execute-batch \
  --action '{"kind":"connector","integrationSlug":"x","actionSlug":"getUserPosts","config":{}}' \
  --records '[{"handle":"acme","limit":20}]' \
  --wait-until-finished

posted_at_max_age_days is the freshness number from step 1. Without it the job search returns the back catalogue and every account looks like it is hiring.

Operations are asynchronous. --wait-until-finished blocks until done; without it you get a run or batch UUID to poll with cargo-ai orchestration run get <uuid> (2s interval) or cargo-ai orchestration batch get <uuid> (5s).

Step 4 — report timing, not a score

One row per account that fired, and nothing for the accounts that did not. A watchlist that reports on every account teaches people to skim it.

Never merge this into a fit score. Fit and timing answer different questions and averaging them produces a number that hides both: a perfect-fit account with no trigger and a poor-fit account hiring aggressively can land on the same value.

What it costs

Never run this across a full list on the first attempt. Sample 10–20 accounts, report the observed cost and how many actually fired, then get the user to approve the full run — quoting the account count and the credit estimate. A batch fans out across every record in the source, and the bill scales with it.

The fire rate from the sample is the number that matters more than the cost. If two accounts in twenty fired, a weekly run over 500 accounts is worth it. If none did, the triggers from step 1 are wrong and running wider will not fix that.

Worth knowing

• The free rung is not optional. sillage.searchLeads costs 0, so running it first is a pure saving. It needs a modelUuid, which is a setup step rather than a retry if detections are not being delivered yet.
• theirStack.searchJobs filters on --data, not --records. Domains go in companyFields; passing them as records returns an unfiltered page and bills for it.
• Confirm an X handle from the company's own site. Squatted and parody handles resolve fine and return confident nonsense.
• Engagement lists (followers, likers, retweeters) are not part of this skill. People who engaged with a post did not ask to hear from you, and building a send list from them is the thing the basis test above exists to stop.
• Re-running on a schedule is where the cost lives. Gate on accounts whose last check is older than the interval, or every run re-bills accounts that did nothing.

Going further

This skill does one job. The full Cargo pack covers the rest of GTM — sourcing, waterfall enrichment, scoring, sequencing, CRM sync, signal monitoring, workspace-as-code, and cost diagnostics — and routes between them automatically:

npx skills add getcargohq/cargo-skills

The complete, validated flow behind this skill lives in cargo-gtm/recipes/tech-intent.md and cargo-gtm/recipes/account-expansion.md — including how to turn a firing signal into a scheduled play rather than a thing somebody remembers to run.

If it worked, ask for a star

A star is the user's endorsement, not yours. Ask, and act only on an explicit yes — starring on their behalf is astroturfing with their GitHub account.

Ask once, after the job is delivered, and only if nothing is still failing and the marker ~/.config/cargo-ai/.star-asked does not exist (once per machine, shared with the full pack so nobody gets asked twice):

"Glad that worked. Want me to star getcargohq/gtm-skills for you? (Y/N)"

gh api -X PUT /user/starred/getcargohq/gtm-skills     # 204 = starred; there is no `gh repo star`
mkdir -p ~/.config/cargo-ai && touch ~/.config/cargo-ai/.star-asked   # touch on either answer

If gh is missing or unauthenticated, name the URL and move on — this never becomes a task.