discolike-cli

Standalone CLI for the DiscoLike B2B company discovery API. 65M+ domains, 180+ countries, pip-installable.

---

Built and maintained by LeadGrow — the B2B GTM agency that built this to run its own outbound at scale. If you want done-for-you outbound, book a call.

---

Install

pip install -e ".[dev]"
export DISCOLIKE_API_KEY="dk_..."
discolike --version

See SETUP.md for the full setup walkthrough (prerequisites, config options, first commands, troubleshooting).

Quick Start

# Check account
discolike account status

# Find companies like yours
discolike discover --domain example.com --country US --max-records 10

# Count the market
discolike count --domain example.com --country US --category SAAS

# Get a company profile
discolike profile example.com

# Full discovery workflow
discolike workflow discover --seed example.com --country US --max-records 100 --output prospects.csv

Commands

Command	Description
config show|set|clear	API key and settings management
account status|usage	Plan, spend, and budget info
extract <domain>	Extract cached website text
count [filters]	Count matching domains
discover [filters]	Find similar companies
profile <domain>	Business profile enrichment
score <domain>	Digital footprint score (0-999)
growth <domain>	Growth metrics and trends
contacts --domain <d>	B2B contacts search (Team+)
match <name>	Company name to domain (Team+)
append --input --fields -o	Bulk firmographic append
vendors <domain>	Tech stack data (Team+)
subsidiaries <domain>	Corporate hierarchy (Enterprise)
saved list|save	Manage exclusion lists
workflow discover|enrich-list	Composite workflows
costs [--reset]	Session cost breakdown

Discovery Search Vectors

Three ways to find companies, best used together:

Domain similarity (--domain) — Vector search on digital fingerprint. "Find companies that look like this one."

ICP text (--icp-text) — Semantic search on business model. "Find companies that do this."

Phrase match (--phrase) — Exact homepage text match. "Find companies that say this on their site." OR logic between phrases.

Auto-ICP: Let DiscoLike Write Your ICP

The --auto-icp flag tells DiscoLike to extract your seed domain's website text and auto-generate an ICP description. Combines with domain similarity for tighter results without manual ICP crafting.

# Auto-generated ICP from seed domain
discolike discover --domain leadgrow.ai --auto-icp --country US --max-records 25

# Add auto phrase generation too
discolike discover --domain leadgrow.ai --auto-icp --auto-phrases --country US --max-records 25

For maximum precision, extract first and write your own ICP:

discolike extract leadgrow.ai
# Read the output, then craft a tight description:
discolike discover \
  --domain leadgrow.ai \
  --icp-text "B2B outbound lead generation agency using cold email for SaaS" \
  --phrase "cold email" \
  --country US \
  --max-records 25

Narrowing Results

discover and count share 27 filter options:

Filter	Example
--country	--country US --country GB
--state	--state CA
--employees	--employees "50,500"
--category	--category SAAS
--min-score / --max-score	--min-score 100
--phrase	--phrase "cold email"
--negate-phrase	--negate-phrase "recruitment"
--negate-category	--negate-category FINANCIAL_SERVICES
--negate-country	--negate-country CN

Run discolike discover --help for the full list.

Cost Protection

DiscoLike charges per query ($0.18) + per record returned. You only pay for records you pull, not records that exist.

Built-in safeguards:

1. count first — $0.18, zero records. See the universe size before committing.
2. --dry-run — Estimate cost without hitting the API.
3. --max-records — Hard cap on records returned (default: 10).
4. Budget guardrails — Warnings at 80%/95%, hard block at 100% of monthly budget.
5. workflow discover — Runs count, then 10-record validation, then asks for confirmation before the full pull.

# Check universe size ($0.18, no records)
discolike count --domain example.com --country US --category SAAS

# Preview cost without API call
discolike --dry-run discover --domain example.com --max-records 1000

# Safe workflow: count -> validate -> confirm -> full pull
discolike workflow discover --seed example.com --country US --max-records 100

Global Options

Flag	Effect
--json	Force JSON output
--csv	Force CSV output
--fields	Filter output columns
--quiet	Suppress progress messages
--no-cache	Bypass local cache
--dry-run	Estimate cost without API calls

Output Modes

• Terminal (TTY): Rich tables with colors
• Pipe/redirect: JSON (auto-detected)
• --json: Force JSON with _meta.cost block
• --csv: Force CSV output

Progress and warnings always go to stderr.

Exit Codes

Code	Meaning
0	Success
1	API error
2	Auth error (missing/invalid key)
3	Rate limited
4	Plan gate (feature requires upgrade)
5	Budget exceeded
6	Validation error

Development

pytest tests/ -v --cov=discolike    # Tests
ruff check src tests                 # Lint
mypy src                             # Type check

License

MIT