Systems Lab

Agent skill

gtm-account-research-brief

Write a one-page research brief on a target account before outreach or a call.

activeReaches the webInstructions only735 words

Filed under Prospecting and list building.

From loootai/gtm-skills · 15 skill entries · 0 · pushed 2026-10-08

What it does when it runs

Write a one-page research brief on a target account before outreach or a call. Use when the user says "research acme.com before my call", "account brief for these 5 companies", "what should I know about this prospect", "prep me for a meeting with Acme" or "pre-call research". Pulls firmographics, LinkedIn page, tech stack, open roles, recent news and the homepage through looot, about $0.02 to $0.05 per account.

Automated analysis of the skill and the 0 files bundled beside it. A skill’s own description is written to be selected by an agent, so it describes the job and not the dependencies.

Keys and connectors you must supply
None found.
Hosts it reaches
  • api.looot.ai
Tool permissions it declares
No allowed-tools in the frontmatter. It only issues instructions, so there is nothing to bound.
Actions present in the files
None. Instructions only.

Ask about gtm-account-research-brief

Opens your assistant with this page's verified links already in the prompt.

Is this safe to install?ClaudeChatGPT
Adapt it to my stackClaudeChatGPT
What else do I need for it to workClaudeChatGPT
Rather ask a human? Talk to Cheetah
git clone --depth 1 --filter=blob:none --sparse https://github.com/loootai/gtm-skills.git /tmp/gtm-skills
git -C /tmp/gtm-skills sparse-checkout set "skills/gtm-account-research-brief"
mkdir -p ~/.claude/skills/gtm-account-research-brief
cp -R "/tmp/gtm-skills/skills/gtm-account-research-brief/." ~/.claude/skills/gtm-account-research-brief/

Picked up without a restart. A project skill of the same name is shadowed by your personal one. For one repository only, swap ~/.claude/skills for .claude/skills. Claude Code docs ↗

The folder is the same in every client that implements the format — 46 of them — so if yours is not above, only the destination changes.

Reproduced in full from loootai/gtm-skills/blob/1cb22eede6829b8f9fa87a78e6957a97e4032bd5/skills/gtm-account-research-brief/SKILL.md, which is licensed MIT (repository). 735 words, 14 headings.

Account research brief

Turns a company domain into a one-page brief a rep can read in two minutes: what the company does, size and funding, what it runs on, who it is hiring, what changed recently, and three angles for the first message. Every fact carries its source.

Needs looot connected (MCP server https://api.looot.ai/mcp, or the looot skill from npx skills add loootai/looot-skills). Searching the catalog and inspecting are free. Runs are paid from the prepaid balance.

Ask for these inputs

  • One or more company domains (acme.com). A company name works for step 1, but a domain is more reliable.
  • What the user sells, in one sentence. The angles in the brief depend on it.
  • Optional: the person they are meeting, with a LinkedIn URL.

Steps and prices

Prices are the cheapest live provider in the public looot catalog on 2026-10-08 (GET https://api.looot.ai/v1/catalog/overview?depth=jobs). The provider looot routes to can cost more. inspect shows the live price and the most a run can hold.

#Job idInputPrice on 2026-10-08
1company.enrichdomainfrom $0.0019 per call
2linkedin.company.profilelinkedin_url from step 1, or domainfrom $0.0015 per call
3company.technographicsdomainfrom $0.01 per result
4jobs.searchthe company name, plus role words if usefulfrom $0.0005 per call
5news.search"<company name>" as queryfrom $0.00099 per call
6web.scrape.markdownurl of the homepage, then /pricing if it existsfrom $0.0002 per call
7 (optional)linkedin.person.profilethe person's linkedin_urlfrom $0.0012 per call

Expected total: about $0.02 per account with the cheapest providers, up to about $0.05 when a step falls back. Steps 3 to 6 do not depend on each other and can run in parallel.

Quote before you run

  1. Count the accounts and the steps you will run. Multiply by the per-step prices above (or by costPerSuccessUsd from search_catalog).
  2. Tell the user the total and the current balance, then wait for a yes. Above 100 runs in one go, always wait for an explicit yes.
  3. Put a cap on every run with fallback.maxCostUsd, for example 0.05. A provider priced above the cap is skipped and nothing is charged for it.

Run it

Run each step as a job with fallback and a fresh idempotency key per account and step (<domain>-enrich, <domain>-news):

run {"endpointId": "job:company.enrich", "input": {"domain": "acme.com"}, "idempotencyKey": "brief-acme.com-enrich", "fallback": {"maxAttempts": 3, "maxCostUsd": 0.05}}
run {"endpointId": "job:news.search", "input": {"query": "\"Acme Inc\""}, "idempotencyKey": "brief-acme.com-news", "fallback": {"maxAttempts": 3, "maxCostUsd": 0.02}}

For jobs.search and company.technographics, call inspect once first: input names differ between providers, and jobInputs in a search_catalog answer lists the ones each job takes.

Read status and error of every run, not only the HTTP status. A miss is a normal answer: write "not found" in the brief, do not guess.

Output format

One markdown block per account:

# Acme Inc (acme.com)
Sources: company.enrich via <provider>, news.search via <provider>, ... | cost $0.023

## Snapshot
- What they do: <one sentence from the homepage, quoted where possible>
- Size: <employees> | HQ: <city, country> | Founded: <year>
- Funding: <last round, amount, date> or "not found"

## Stack
<up to 10 technologies, grouped: CRM, marketing, data, cloud>

## Hiring
<count of open roles, the 3 most relevant titles with links>

## Recent news (last 90 days)
- <date> <headline> (<source link>)

## Angles for the first message
1. <angle tied to one fact above, with the fact named>
2. ...
3. ...

Rules for the angles: each one points at a fact in the brief. No invented pain points, no flattery lines.

Cost and stop rules

  • Never loop a failing call. Two failures on the same step: skip the step, mark it "not found" and say so.
  • Report the total actualCost at the end, per account and overall.
  • Do not run step 3 (per-result pricing) on more than 20 accounts without a fresh quote.

When not to use this skill

  • One fact only (headcount, email format): use the single-job skills in loootai/looot-skills.
  • A list of many accounts to qualify against an ICP: use gtm-icp-lookalike-list or gtm-tech-stack-targeting.

Other skills for the same job

Different authors, same problem. Matched on the words in the skill name, across every library in the catalogue except this one.

Need help setting it up?

This page tells you what gtm-account-research-brief does and what it needs. Cheetah builds the agent setup it runs inside: data, CRM, sequencing and the guardrails.

Book a call →

The directory stays free. There is nothing gated behind this.