Moltsets

Agentic TAM mapping from natural language.

Moltsets lets you type the companies you want to find and returns campaign-ready account sets with real domains, segment reasoning, usage tracking, exports, and next-step people-search requests.

npm install -g truetolife
moltsets doctor
moltsets

The npm package is only truetolife. It installs three command aliases:

truetolife doctor
ttlt doctor
moltsets doctor

The CLI expects the local taxonomy-demo MCP credentials to already be configured. Send the MCP credentials separately from the npm install command.

Why Moltsets

Most data tools give you lists. Moltsets gives you the audience story, the account set, and the next workflow.

Instead of filtering by generic firmographics, you can ask for a specific business situation:

I sell industrial account intelligence software. Find specialty chemical companies that produce coatings, adhesives, performance materials, polymers, additives, water treatment products, personal care ingredients, or sustainability-focused industrial materials.

Moltsets uses the TAM Engine to form a mapped account set, explain why the accounts match, and ask what you want to do next.

Quickstart

npm install -g truetolife
moltsets doctor
moltsets

Or pass a prompt directly:

moltsets query "I sell senior living referral CRM. Find senior living operators with memory care, assisted living, independent living, and retirement community positioning."

Export immediately:

moltsets query "Find aerospace companies involved in avionics, defense aviation, aftermarket support, MRO, and satcom." --format csv

Example Output

Moltset formed: industrial account intelligence
Moltsets mapped this request to 31 accounts using specialty chemicals, technical applications.

Name        URL                         Category                                  Signals
altana      https://www.altana.com       Products & Manufacturing > Specialty...   coatings, sustainability
arxada      https://www.arxada.com       Products & Manufacturing > Specialty...   adhesives, water treatment
chemours    https://www.chemours.com     Products & Manufacturing > Specialty...   performance materials

Usage

| Plan | Price | Records / 5 Hours | Weekly Record Cap | Best For | |---|---:|---:|---:|---| | $97 Growth | $97/mo | 5,000 | 35,000 | Serious individual agents, small Clay + Claude workflows |

Usage is based on companies returned. For v1, usage tracking is local and demo-grade.

Usage
5h records   [████████░░░░░░░░░░░░] 2,000 / 5,000
weekly       [██████░░░░░░░░░░░░░░] 11,250 / 35,000

Commands

truetolife                   # guided mode
ttlt                         # guided mode
moltsets                     # guided mode
moltsets query "..."         # run a natural-language TAM request
moltsets demo                # choose a live example prompt
moltsets usage               # show usage bars
moltsets usage --json        # show usage JSON
moltsets config              # show config paths
moltsets doctor              # verify local taxonomy MCP credentials

Exports

Moltsets supports:

• terminal table
• CSV
• JSON
• Markdown report
• generic JSON webhook

moltsets query "Find community banks with treasury and digital banking" --format markdown
moltsets query "Find mortgage lenders with refinance and first-time buyer journeys" --webhook https://example.com/webhook

People Search Step

After forming a Moltset, the CLI asks whether you want to find people at those companies. It guides you toward personas such as directors, VPs, department heads, operations leaders, finance leaders, or technical/product leaders.

In v1, the CLI prepares a structured people-search request object. A people provider can be connected later without changing the main TAM mapping flow.

Configuration

Packaged defaults:

config/followups.json
config/demo-queries.json
flows/guided.md

User overrides:

~/.moltsets/config/followups.json
~/.moltsets/flows/guided.md

If a user override is invalid, Moltsets warns and falls back to packaged defaults.

TAM Engine

The CLI calls a hidden TAM Engine adapter. Backend details are intentionally not exposed in user-facing output.

For this demo, every recipient needs local MCP credentials. The normal path is:

1. Install the CLI from npm.
2. Add the taxonomy-demo MCP credentials you were given to the local Codex config.
3. Run moltsets doctor.
4. Run moltsets.

Example config shape:

[mcp_servers.taxonomy-demo]
url = "https://YOUR-MCP-ENDPOINT/mcp"

[mcp_servers.taxonomy-demo.headers]
Authorization = "Bearer YOUR_TOKEN"

Moltsets reads this from:

~/.codex/config.toml

You can also use environment variables:

export MOLTSETS_TAM_ENGINE_URL="https://..."
export MOLTSETS_TAM_ENGINE_TOKEN="..."

Development

npm install
npm run build
npm test
npm install -g .
moltsets usage

Roadmap

• Hosted usage enforcement
• Fully live people-search provider
• Saved Moltset reopen/re-run commands
• More export destinations
• Team workspaces

License

MIT