Agent skill
account-scoring
Build a target account list from the ICP in strategy.json with DiscoLike only, qualify each account yes/no with DiscoLike's ICP check, score it 0-100 and write qualified_accounts.csv with score and reason.
Filed under Prospecting and list building.
From sccsadmin11/outbound-gtm-sc · 11 skill entries · 0 · pushed 2026-10-22
What it does when it runs
Build a target account list from the ICP in strategy.json with DiscoLike only, qualify each account yes/no with DiscoLike's ICP check, score it 0-100 and write qualified_accounts.csv with score and reason.
Automated analysis of the skill and the 3 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
- DISCOLIKE_API_KEY
- Hosts it reaches
- api.discolike.com
- discolike.com
- docs.discolike.com
- www.linkedin.com
- Tool permissions it declares
- No
allowed-toolsin the frontmatter. It does act, so it runs under whatever permissions your session already grants. - Actions present in the files
- shell
Ask about account-scoring
Opens your assistant with this page's recorded context. Check the linked sources before relying on it.
Install it
View source on GitHub ↗git clone --depth 1 --filter=blob:none --sparse https://github.com/sccsadmin11/outbound-gtm-sc.git /tmp/outbound-gtm-sc git -C /tmp/outbound-gtm-sc sparse-checkout set "03-account-scoring" mkdir -p ~/.claude/skills/account-scoring cp -R "/tmp/outbound-gtm-sc/03-account-scoring/." ~/.claude/skills/account-scoring/
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.
Before you install: this skill will not complete its job on a bare agent. It needs DISCOLIKE_API_KEY, which you have to obtain separately.
The skill
Source on GitHub ↗Reproduced in full from sccsadmin11/outbound-gtm-sc/blob/fc996416957ff6232cc00c3b5c94d601b68e1d61/03-account-scoring/SKILL.md, which is licensed MIT (repository). 342 words, 10 headings.
03 Account scoring (DiscoLike only)
Purpose
Turn the ICP into a ranked list of companies that actually fit, with a short reason for every yes and no. This skill uses DiscoLike and nothing else.
Inputs
strategy.jsonfrom skill 02 (icp.icp_text,negate_icp_text,countries,employee_range,lookalike_domains,exclude).
Outputs
qualified_accounts.csv: only qualified accounts, sorted by score.scored_accounts.csv: every account with score and reason (useful to tune the ICP).
Columns: domain, company_name, qualified, score, reason, similarity, icp_fit, icp_confidence, footprint, employees, industry, linkedin_url, description.
Env vars
| Var | Where to get it |
|---|---|
DISCOLIKE_API_KEY | DiscoLike app -> Account -> API keys (https://discolike.com) |
How it works
GET /v1/discoverwithicp_text,negate_icp_text,country,employee_range, lookalikedomains,exclude_domain(your own domain is always excluded) andexclude_leadgen=true.POST /v1/validate/icpwithintegration_id: "native-icp"on the returned domains, then pollGET /v1/discogen/status/{task_id}. Each domain gets Fit yes/no, a confidence and a reasoning line.- Score 0-100 = 55% similarity + 35% ICP confidence + 10% digital footprint (capped at 400 of 800). Qualified = Fit yes AND score >=
--min-score(default 60).
Steps (agent run instructions)
- Preview the query, no API call:
python 03-account-scoring/scripts/score_accounts.py --strategy strategy.json --dry-run - Check the market size:
python 03-account-scoring/scripts/score_accounts.py --strategy strategy.json --count-only - Run it (start small, then raise
--max-records):python 03-account-scoring/scripts/score_accounts.py --strategy strategy.json --max-records 200 --min-score 60 - Read 10 rows of
scored_accounts.csvwith the user. If good companies are scored "no", tightenicp_textornegate_icp_textinstrategy.jsonand re-run.
Notes
- DiscoLike runs at most 3 ICP validation tasks at once. The script waits and retries on 429.
--skip-validatescores on similarity + footprint only. Cheaper, less accurate.
Open TODOs
- The
/discogen/statusresponse schema is not published. The parser readsresultsas a dict or list and finds the Fit / Confidence / Reasoning columns by name, case-insensitive. - Max domains per
/validate/icpcall is not documented. The script sends batches of 100.
Handoff
qualified_accounts.csv -> 04-people-waterfall.
Files bundled with it
These load only when the skill asks for them, so they cost nothing until it runs.
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.
- account-structure-review by irinabuht12-oss · 6,620
- weekly-account-summary by irinabuht12-oss · 6,620
- positive-reply-scoring by growthenginenowoslawski · 797
- lead-scoring by shawnpang · 374
- find-lead-account-owner by zapier · 354
- ai-audit-account-report by zapier · 354
- named-account-trigger-radar by zapier · 354
- account-contact-shortlist by explorium-ai · 348
Need help setting it up?
This page tells you what account-scoring 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.