Skip to content

getaero-io/gtm-eng-skills / skills/deepline-gtm

deepline-gtm

Unverified

GTM leads, enrichment, outreach, Plays, developer signals/Reo.dev. Providers: adyntel,affinity,ai_ark,akta,allegrow,amplemarket,apify,attention,attio,aviato,bettercontact,bigquery,bluesky,bounceban,brave,browserbase,builtwith,buyercaddy,cbinsights,clay,clickhouse,close,cloudflare,contactout,contextdev,crunchbase,crustdata-v3,customer_db,databricks,dataforseo,datagma,dealroom,deepline_ip_to_company,deepline_native,deeplineagent,deepseek,discolike,dnb,edges,emailbison,emailguard,enformion,enigma,exa,findymail,firecrawl,fireflies,firmable,forager,fullenrich,fundable,generic_http,getleads,gong,google_ads_audiences,google_workspace,govfiles,grain,hackernews,harvestapi,heyreach,hubspot,hunter,icypeas,instantly,intercom,ipqs,kaspr,kernel,kickbox,leadmagic,lemlist,limadata,linkedin_ads_audiences,luma,lusha,meerapi,meta_audiences,millionverifier,nooks,notion,openmart,opensosdata,openwebninja,outreach,parallel,particle,peopledatalabs,pitchbook,podscan,polymarket,postgres,predictleads,prospeo,quickenrich,rapidapi,….

FileKindSize
SKILL.mdskill md15.9 KB
claude-deepline-statusline.mjsexecutablescript25.0 KB
prompts.jsonconfig297.9 KB
provider-playbooks/adyntel.mdmarkdown1.2 KB
provider-playbooks/affinity.mdmarkdown825 B
provider-playbooks/ai-ark.mdmarkdown10.6 KB
provider-playbooks/akta.mdmarkdown1.5 KB
provider-playbooks/allegrow.mdmarkdown4.6 KB
provider-playbooks/amplemarket.mdmarkdown2.8 KB
provider-playbooks/apify.mdmarkdown1.8 KB
provider-playbooks/attention.mdmarkdown449 B
provider-playbooks/attio.mdmarkdown5.6 KB
provider-playbooks/aviato.mdmarkdown756 B
provider-playbooks/bettercontact.mdmarkdown1.7 KB
provider-playbooks/bigquery.mdmarkdown1019 B
provider-playbooks/bloomberry.mdmarkdown935 B
provider-playbooks/bluesky.mdmarkdown139 B
provider-playbooks/bounceban.mdmarkdown1.6 KB
provider-playbooks/brave.mdmarkdown422 B
provider-playbooks/browserbase.mdmarkdown1.0 KB
provider-playbooks/builtwith.mdmarkdown1.0 KB
provider-playbooks/buyercaddy.mdmarkdown2.3 KB
provider-playbooks/cbinsights.mdmarkdown436 B
provider-playbooks/clay.mdmarkdown875 B
provider-playbooks/clickhouse.mdmarkdown1.1 KB
provider-playbooks/close.mdmarkdown2.6 KB
provider-playbooks/cloudflare.mdmarkdown2.3 KB
provider-playbooks/contactout.mdmarkdown8.3 KB
provider-playbooks/contextdev.mdmarkdown582 B
provider-playbooks/crunchbase.mdmarkdown742 B
provider-playbooks/crustdata-v2.mdmarkdown490 B
provider-playbooks/crustdata-v3.mdmarkdown8.4 KB
provider-playbooks/crustdata.mdmarkdown6.8 KB
provider-playbooks/databricks.mdmarkdown3.5 KB
provider-playbooks/dataforseo.mdmarkdown950 B
provider-playbooks/datagma.mdmarkdown6.0 KB
provider-playbooks/dealroom.mdmarkdown632 B
provider-playbooks/deepline-ip-to-company.mdmarkdown480 B
provider-playbooks/deepline-native-osint.mdmarkdown1.5 KB
provider-playbooks/deepline-native.mdmarkdown25.1 KB
provider-playbooks/deeplineagent.mdmarkdown2.7 KB
provider-playbooks/deepseek.mdmarkdown1.2 KB
provider-playbooks/discolike.mdmarkdown6.2 KB
provider-playbooks/dnb.mdmarkdown1.4 KB
provider-playbooks/dropleads.mdmarkdown10.2 KB
provider-playbooks/edges.mdmarkdown875 B
provider-playbooks/emailbison.mdmarkdown634 B
provider-playbooks/emailguard.mdmarkdown243 B
provider-playbooks/enformion.mdmarkdown9.3 KB
provider-playbooks/enigma.mdmarkdown1.7 KB
provider-playbooks/exa.mdmarkdown1.4 KB
provider-playbooks/file-transport.mdmarkdown739 B
provider-playbooks/findymail.mdmarkdown880 B
provider-playbooks/firecrawl.mdmarkdown2.4 KB
provider-playbooks/fireflies.mdmarkdown988 B
provider-playbooks/firmable.mdmarkdown1.3 KB
provider-playbooks/forager.mdmarkdown2.3 KB
provider-playbooks/fullenrich.mdmarkdown2.1 KB
provider-playbooks/fundable.mdmarkdown2.0 KB
provider-playbooks/generic-http.mdmarkdown348 B
provider-playbooks/getleads.mdmarkdown2.2 KB
provider-playbooks/gong.mdmarkdown488 B
provider-playbooks/google-ads-audiences.mdmarkdown2.0 KB
provider-playbooks/google-calendar-availability.mdmarkdown1.6 KB
provider-playbooks/google-calendar.mdmarkdown1.1 KB
provider-playbooks/google-sheets.mdmarkdown842 B
provider-playbooks/google-workspace.mdmarkdown2.1 KB
provider-playbooks/govfiles.mdmarkdown1.8 KB
provider-playbooks/grain.mdmarkdown1.0 KB
provider-playbooks/hackernews.mdmarkdown208 B
provider-playbooks/harvestapi.mdmarkdown3.5 KB
provider-playbooks/heyreach.mdmarkdown1.6 KB
provider-playbooks/hubspot.mdmarkdown3.3 KB
provider-playbooks/hunter.mdmarkdown849 B
provider-playbooks/icypeas.mdmarkdown1.8 KB
provider-playbooks/instantly.mdmarkdown1.6 KB
provider-playbooks/integration-operation.mdmarkdown704 B
provider-playbooks/intercom.mdmarkdown275 B
provider-playbooks/ipqs.mdmarkdown1.7 KB
provider-playbooks/kaspr.mdmarkdown822 B
provider-playbooks/kernel.mdmarkdown875 B
provider-playbooks/leadmagic.mdmarkdown5.3 KB
provider-playbooks/lemlist.mdmarkdown7.5 KB
provider-playbooks/limadata.mdmarkdown636 B
provider-playbooks/linkedin-ads-audiences.mdmarkdown629 B
provider-playbooks/luma.mdmarkdown3.2 KB
provider-playbooks/lusha.mdmarkdown3.5 KB
provider-playbooks/meerapi.mdmarkdown1.2 KB
provider-playbooks/meta-audiences.mdmarkdown4.8 KB
provider-playbooks/millionverifier.mdmarkdown618 B
provider-playbooks/nooks.mdmarkdown1.0 KB
provider-playbooks/notion.mdmarkdown724 B
provider-playbooks/openmart.mdmarkdown2.3 KB
provider-playbooks/opensosdata.mdmarkdown4.0 KB
provider-playbooks/openwebninja.mdmarkdown315 B
provider-playbooks/outreach.mdmarkdown2.6 KB
provider-playbooks/parallel.mdmarkdown4.4 KB
provider-playbooks/particle.mdmarkdown2.0 KB
provider-playbooks/peopledatalabs.mdmarkdown4.5 KB
provider-playbooks/pipedrive.mdmarkdown582 B
provider-playbooks/pitchbook.mdmarkdown606 B
provider-playbooks/podscan.mdmarkdown2.7 KB
provider-playbooks/polymarket.mdmarkdown2.2 KB
provider-playbooks/postgres.mdmarkdown553 B
provider-playbooks/predictleads.mdmarkdown622 B
provider-playbooks/prospeo.mdmarkdown2.4 KB
provider-playbooks/quickenrich.mdmarkdown595 B
provider-playbooks/rapidapi.mdmarkdown2.5 KB
provider-playbooks/rb2b.mdmarkdown247 B
provider-playbooks/redshift.mdmarkdown763 B
provider-playbooks/reo-write.mdmarkdown1.6 KB
provider-playbooks/reo.mdmarkdown2.0 KB
provider-playbooks/resend.mdmarkdown3.6 KB
provider-playbooks/salesforce.mdmarkdown3.6 KB
provider-playbooks/salesforge.mdmarkdown274 B
provider-playbooks/salesloft.mdmarkdown1.0 KB
provider-playbooks/scrapecreators.mdmarkdown3.4 KB
provider-playbooks/searchbug.mdmarkdown637 B
provider-playbooks/sec-edgar.mdmarkdown1.6 KB
provider-playbooks/sentrion.mdmarkdown737 B
provider-playbooks/serper.mdmarkdown1011 B
provider-playbooks/slack.mdmarkdown21.1 KB
provider-playbooks/smartleads.mdmarkdown11.8 KB
provider-playbooks/snitcher.mdmarkdown302 B
provider-playbooks/snowflake.mdmarkdown768 B
provider-playbooks/socrata.mdmarkdown3.2 KB
provider-playbooks/sumble.mdmarkdown1.3 KB
provider-playbooks/tamradar.mdmarkdown1.4 KB
provider-playbooks/test-unhinted.mdmarkdown505 B
provider-playbooks/test.mdmarkdown254 B
provider-playbooks/theirstack.mdmarkdown4.4 KB
provider-playbooks/theswarm.mdmarkdown1.4 KB
provider-playbooks/treg.mdmarkdown1.5 KB
provider-playbooks/trestle.mdmarkdown3.7 KB
provider-playbooks/trykitt.mdmarkdown862 B
provider-playbooks/twitterapi.mdmarkdown3.6 KB
provider-playbooks/unipile.mdmarkdown4.8 KB
provider-playbooks/upcell.mdmarkdown715 B
provider-playbooks/vector.mdmarkdown257 B
provider-playbooks/versium.mdmarkdown690 B
provider-playbooks/wiza.mdmarkdown2.0 KB
provider-playbooks/wizleads.mdmarkdown1.1 KB
provider-playbooks/zerobounce.mdmarkdown1.3 KB
provider-playbooks/zoho-crm.mdmarkdown630 B
provider-playbooks/zoominfo.mdmarkdown1.2 KB
recipes/account-orgchart.mdmarkdown35.6 KB
recipes/build-tam.mdmarkdown3.7 KB
recipes/clay-to-deepline.mdmarkdown26.8 KB
recipes/deepline-monitors.mdmarkdown32.5 KB
recipes/developer-signals.mdmarkdown14.7 KB
recipes/find-qualified-titles.mdmarkdown5.2 KB
recipes/portfolio-prospecting.mdmarkdown3.6 KB
recipes/small-business-prospecting.mdmarkdown3.0 KB
references/clay-action-mappings.mdmarkdown40.0 KB
references/clay-api-surface.mdmarkdown18.6 KB
references/clay-extraction.mdmarkdown16.2 KB
references/industry-radars.mdmarkdown3.3 KB
references/monitor-contract-reference.mdmarkdown25.9 KB
references/monitor-operations.mdmarkdown12.4 KB
references/monitor-sdk.mdmarkdown3.2 KB
references/run-events.mdmarkdown14.3 KB
scripts/clay-extract-bookmarklet.jsscript10.5 KB
scripts/clay-extract-bookmarklet.url.txtmarkdown8.2 KB
scripts/clay-extract.pyscript17.9 KB
scripts/clay-har-miner.pyscript5.8 KB
scripts/clay-to-deepline.mjsscript134.1 KB
scripts/contact-accuracy-audit.pyscript8.9 KB
scripts/fixtures_contact_accuracy_audit.jsonconfig5.5 KB
scripts/fixtures_current_role.jsonconfig10.3 KB
scripts/flatten-search-contact-persons.pyscript3.3 KB
scripts/select-current-role.pyscript12.2 KB
scripts/validate-emails.pyexecutablescript2.0 KB
skill-metadata.jsonconfig2.2 KB
writing-outreach.mdmarkdown14.1 KB

references/clay-action-mappings.md

View on GitHub
# Clay Action → Deepline Tool Mappings

<!-- skill-tool-compat: run_javascript -- Supported local runtime tool retained in explicitly historical payload examples; author new deterministic transforms as native Play callbacks. -->
<!-- skill-tool-field: clay_record -- Reserved column alias for the extracted Clay record, not a provider tool. -->

Every Clay action maps to a specific Deepline CLI tool or native play. Use actual tool IDs in every generated script — never generic descriptions.

## Legacy payload reference — not runnable Play syntax

The string-form `.withColumn(alias, tool, payload)` snippets below retain
historical payloads, prompts, field paths, provider order, and validation
semantics. They are **not** current callback-based Play examples. Do not paste
them into a Play.
Translate a selected recipe using `deepline plays --help`:
`withColumn(alias, async (row) => ...)`, `deepline.tools.execute` with
stable `id`, `tool`, typed `input`, and `description`, or `deepline.runPlay` for a
prebuilt. Replace brace interpolation with typed row reads and use declared
getters rather than assuming historical `.output.body` wrappers still exist.

Unmigrated blocks remain below for person enrichment, cross-table filesystem
joins, email waterfalls/validation/permutations, company-source Python passes,
normalization, classification/AI/Octave prompts, campaign writes, and field
interpolation. Their output getters, staging, and side-effect semantics need
case-by-case contract validation before conversion; no faithful runnable
replacement is claimed here. Retain their payload knowledge, but use
[build, run and retrieve](../SKILL.md) in the GTM skill for authority and
retrieval. Campaign writes require explicit authority; a provider miss does not
authorize adding the fallback routes listed here.

## ⚠️ Tool Discovery Protocol — Read First

**This mapping is a starting-point reference, not a guarantee.** Deepline adds new tools, native plays, and provider integrations continuously. The right mental model:

**For every Clay action, the selection order is:**

1. **Native play first** — check if a native Deepline play covers the action (they're stable, multi-provider, and cost-optimized). Current native plays: `name-and-domain-to-email-waterfall`, `company-to-contact`, and `person-to-phone`.
2. **Search for a dedicated tool** — `deepline search "<input you have -> result you need>"` before hardcoding any individual provider tool. New tools are added regularly. Examples: `deepline search "person profile -> ICP qualification"`, `deepline search "email -> validation status"`, `deepline search "lead list -> leads added to a campaign"`.
3. **Verify the tool exists** - `deepline describe <tool_id>`. If it errors, the tool doesn't exist yet - use the `deeplineagent` fallback from this doc.
4. **Use the mapping below as a fallback** — when no native play or dedicated tool exists.

```bash
# Standard discovery pattern before writing any play column
deepline search "<input you have -> result you need>"     # find current options
deepline describe <candidate_tool_id> # verify it exists + see payload schema
```

**Why this matters:** Deepline may add a native tool for a Clay action at any time. Searching first means your script gets the better tool automatically — instead of being locked into a `deeplineagent` approximation.

---

## Model Translation

| Clay model                       | Deepline column params                  | Notes                                            |
| -------------------------------- | --------------------------------------- | ------------------------------------------------ |
| `gpt-4.1`, `claude` (clay-argon) | `"model":"anthropic/claude-sonnet-4.6"` | Complex reasoning, larger structured outputs     |
| `gpt-4.1-mini`                   | `"model":"openai/gpt-5.6-luna"`         | Clay's mid-tier model for claygent-style columns |
| `gpt-4o-mini`, `gpt-5-mini`      | `"model":"openai/gpt-5.6-luna"`         | Fast classify/generate                           |
| `gpt-5-nano`                     | `"model":"openai/gpt-5.6-luna"`         | Cheapest tier approximation                      |

---

## Complete Action → Tool Mapping

| Clay action key                                             | Deepline tool / native play                                                                                                                                                                                                                                                                                                                        | Notes                                                                                         |
| ----------------------------------------------------------- | -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | --------------------------------------------------------------------------------------------- |
| `find-lists-of-companies-with-mixrank-source` (source type) | **Pass 1**: `crustdata_v3_company_search` — filters by location, employee count, industry, funding, and investors. Returns company name, domain, LinkedIn URL, HQ, funding, and firmographic fields. **Pass 2** (optional): `prospeo_enrich_company` — adds `description`, `employee_count`, `industry`, `type`. See Company Source section below. | Use `crustdata_v3_company_search_autocomplete` first for canonical filter values              |
| `enrich-person-with-mixrank-v2`                             | `leadmagic_profile_search` → `crustdata_v3_person_enrich` waterfall                                                                                                                                                                                                                                                                                | See Person Enrichment section                                                                 |
| `lookup-company-in-other-table`                             | `run_javascript` (local CSV join)                                                                                                                                                                                                                                                                                                                  | Export company table to CSV first                                                             |
| `lookup-multiple-rows-in-other-table`                       | `run_javascript` (local CSV join)                                                                                                                                                                                                                                                                                                                  | Same pattern                                                                                  |
| `chat-gpt-schema-mapper`                                    | `deeplineagent`; use `jsonSchema` when you need structured extraction                                                                                                                                                                                                                                                                              | Single-value classification                                                                   |
| `normalize-company-name`                                    | `deeplineagent` or `run_javascript`                                                                                                                                                                                                                                                                                                                | JS preferred for pure string ops                                                              |
| `generate-email-permutations`                               | `run_javascript`                                                                                                                                                                                                                                                                                                                                   | Pure compute, no provider                                                                     |
| `validate-email` (all instances)                            | `leadmagic_email_validation` (one final gate)                                                                                                                                                                                                                                                                                                      | Skip per-step validation; validate once after waterfall                                       |
| `wiza-find-email`                                           | `name-and-domain-to-email-waterfall`                                                                                                                                                                                                                                                                                                               | Part of native play                                                                           |
| `find-email-v2` (Hunter)                                    | `hunter_email_finder`                                                                                                                                                                                                                                                                                                                              | Part of native play                                                                           |
| `leadmagic-find-work-email`                                 | `leadmagic_email_finder`                                                                                                                                                                                                                                                                                                                           | Part of native play                                                                           |
| `findymail-find-work-email`                                 | `findymail_find_from_name`                                                                                                                                                                                                                                                                                                                         | Covered by native play                                                                        |
| `enrich-person` (PDL)                                       | `peopledatalabs_enrich_contact` (waterfall step)                                                                                                                                                                                                                                                                                                   | Covered by native play                                                                        |
| `dropcontact-enrich-person`                                 | `name-and-domain-to-email-waterfall`                                                                                                                                                                                                                                                                                                               | Covered by native play                                                                        |
| **Entire email waterfall group**                            | `name-and-domain-to-email-waterfall`                                                                                                                                                                                                                                                                                                               | One play replaces all 6 finders when you have a domain; include `linkedin_url` when available |
| `use-ai` (no web, simple)                                   | `deeplineagent`                                                                                                                                                                                                                                                                                                                                    | Match model tier                                                                              |
| `use-ai` (no web, structured)                               | `deeplineagent` + `jsonSchema`                                                                                                                                                                                                                                                                                                                     |                                                                                               |
| `use-ai` (claygent + web)                                   | Pass 1: `exa_search` → Pass 2: `deeplineagent`                                                                                                                                                                                                                                                                                                     | Always split research and synthesis                                                           |
| `octave-qualify-person`                                     | `deeplineagent`, ICP scoring prompt, `jsonSchema`                                                                                                                                                                                                                                                                                                  | See Octave section                                                                            |
| `octave-enrich-person`                                      | `exa_search` + `deeplineagent`                                                                                                                                                                                                                                                                                                                     |                                                                                               |
| `octave-run-sequence-runner`                                | Pass 1: `deeplineagent` (signals) → Pass 2: `deeplineagent` (email)                                                                                                                                                                                                                                                                                | Always 2 passes                                                                               |
| `social-posts-get-post-activity-posts-and-shares`           | User-provided post export                                                                                                                                                                                                                                                                                                                          | Run-as-button in Clay — omit unless user needs posts                                          |
| `score-your-data` (unconfigured)                            | `run_javascript` keyword scoring                                                                                                                                                                                                                                                                                                                   | See Scoring section                                                                           |
| `add-lead-to-campaign` (Smartlead)                          | `smartlead_push_to_campaign`                                                                                                                                                                                                                                                                                                                       |                                                                                               |
| `add-lead-to-campaign` (Instantly)                          | `instantly_add_to_campaign`                                                                                                                                                                                                                                                                                                                        |                                                                                               |
| `exa_search` (Clay native)                                  | `exa_search`                                                                                                                                                                                                                                                                                                                                       | Direct equivalent                                                                             |

---

## Person Enrichment (LinkedIn URL → profile data)

### `enrich-person-with-mixrank-v2`

**Best single tool** — `leadmagic_profile_search`:

```ts
.withColumn('person_profile', 'leadmagic_profile_search', { profile_url: "{{linkedin_url}}" })
```

Key output paths: `.output.body.full_name`, `.output.body.work_experience[0].company_website`, `.output.body.company_website`

**Richer fallback** — `crustdata_v3_person_enrich`:

```ts
.withColumn('person_profile', 'crustdata_v3_person_enrich', { professional_network_profile_urls: ["{{linkedin_url}}"], fields: ["basic_profile", "experience"] })
```

Key output paths: `.output.body.matches[0].person_data.basic_profile.name`, `.output.body.matches[0].person_data.contact.business_emails[0].email`, `.output.body.matches[0].person_data.experience.employment_details.current[0].company_website_domain`

For work history and posts, use an authorized source or a user-provided export. Do not collect private LinkedIn pages or saved searches without established source-platform permission.

---

## Company Table Lookup (Clay cross-table join)

### `lookup-company-in-other-table` / `lookup-multiple-rows-in-other-table`

Export the linked Clay table to a local CSV first (`clay_fetch_records.sh` schema mode). Then join with `run_javascript`:

```javascript
// $WORKDIR/join_company.js
// Adjust field names to match your table's column aliases
const fs = require('fs');
const rows = fs
  .readFileSync(process.env.COMPANY_CSV_PATH, 'utf8')
  .trim()
  .split('\n')
  .slice(1)
  .map((line) => JSON.parse(line)); // adjust for CSV format
const joinKey = row['company_domain'] || row['domain']; // ← your join key column
return rows.find((c) => c.domain === joinKey) || null;
```

```ts
.withColumn('company_data', 'run_javascript', { code: "@$WORKDIR/join_company.js" })
```

---

## Email Waterfall (6 Clay providers → 1 Deepline native play)

The entire Clay waterfall group collapses to one native play. The play owns its provider order and validation stages. Pick based on available input:

**Have LinkedIn URL + name + domain** (preferred — highest hit rate):

```ts
.withColumn('work_email', 'name-and-domain-to-email-waterfall', { linkedin_url: "{{linkedin_url}}", first_name: "{{first_name}}", last_name: "{{last_name}}", domain: "{{company_domain}}" })

```

**Have name + company only**:

```ts
.withColumn('work_email', 'name-and-domain-to-email-waterfall', { first_name: "{{first_name}}", last_name: "{{last_name}}", domain: "{{company_domain}}" })

```

**Have first + last + domain only** (cost-efficient — tries pattern validation first):

```ts
.withColumn('work_email', 'name-and-domain-to-email-waterfall', { first_name: "{{first_name}}", last_name: "{{last_name}}", domain: "{{domain}}" })

```

**Alternative individual providers** (use `deepline search "name and company domain -> work email"` to see current list):

- `icypeas_email_search` — 700M+ profiles, strong LinkedIn coverage; useful as a step if native play misses
- Run `deepline describe icypeas_email_search` to verify tool exists before using

**Final validation gate** (replaces all per-step Clay `validate-email` calls):

Default — `leadmagic_email_validation`:

```ts
.withColumn('email_valid', 'leadmagic_email_validation', { email: "{{work_email}}" })
```

LeadMagic returns four relevant statuses (as `.output.body.email_status`):

| Status            | Meaning                                            | Bounce rate | Charge   |
| ----------------- | -------------------------------------------------- | ----------- | -------- |
| `valid`           | Verified deliverable                               | <1%         | Yes      |
| `valid_catch_all` | Catch-all domain; engagement data confirms address | <5%         | Yes      |
| `catch_all`       | Domain accepts all; unverifiable                   | Unknown     | **Free** |
| `unknown`         | Mail server no response                            | Unknown     | **Free** |
| `invalid`         | Will bounce                                        | ~100%       | Yes      |

**Accept `valid`, `valid_catch_all`, AND `catch_all` as "found"** — all three are as reliable as what Clay reports. `valid_catch_all` is the highest-confidence version (LeadMagic has engagement signal data for the address). Do not accept `unknown`.

Alternative — `zerobounce_validate` (more detailed sub_status, better for catch-all domains):

```ts
.withColumn('email_valid', 'zerobounce_validate', { email: "{{work_email}}" })
```

Check `.status` and `.sub_status`. Use `zerobounce_validate` for each email; Deepline disables ZeroBounce batch validation because the upstream batch endpoint currently returns 403 Access denied from Deepline egress.

Run `deepline search "email -> validation status"` to see all current options.

---

## Email Permutations

### `generate-email-permutations`

Pure `run_javascript` — no provider:

```javascript
// $WORKDIR/email_permutations.js
const first = (row['first_name'] || '').toLowerCase().replace(/[^a-z]/g, '');
const last = (row['last_name'] || '').toLowerCase().replace(/[^a-z]/g, '');
const domain = row['company_domain'] || '';
if (!first || !last || !domain) return null;
const perms = [
  `${first}.${last}@${domain}`,
  `${first}${last}@${domain}`,
  `${first}_${last}@${domain}`,
  `${first}@${domain}`,
  `${first[0]}${last}@${domain}`,
  `${first}${last[0]}@${domain}`,
  `${first[0]}.${last}@${domain}`,
  `${last}.${first}@${domain}`,
];
return { permutations: perms, comma_separated_list: perms.join(',') };
```

```ts
.withColumn('email_permutations', 'run_javascript', { code: "@$WORKDIR/email_permutations.js" })
```

Prefer `name-and-domain-to-email-waterfall` over a hand-built waterfall when you already have a clean company domain.

---

## Company Source — Replacing `find-lists-of-companies-with-mixrank-source`

Clay's Mixrank source fetches a pre-built list from a configured Mixrank query. The Deepline equivalent is a two-pass Python script: **discover with `crustdata_v3_company_search`**, then **enrich with `prospeo_enrich_company`** for fields Clay gets from Mixrank (description, industry, size, type).

### Pass 1 — Generate company list (`crustdata_v3_company_search`)

```python
import csv, io, json, subprocess

# CrustData filter payload — translate from Clay's Mixrank source config
# Check the Clay table config or ask the user for the original filter criteria
payload = {
    "filters": {
        "op": "and",
        "conditions": [
            {"field": "locations.headquarters", "type": "(.)", "value": "Los Angeles"},
            {"field": "locations.country", "type": "=", "value": "USA"},
            {"field": "taxonomy.categories", "type": "(.)", "value": "software"},
            {"field": "basic_info.employee_count_range", "type": "in", "value": ["51-200", "201-500", "501-1000"]},
        ],
    },
    "limit": 100,
}

payload["fields"] = ["basic_info", "locations"]  # response groups to return

columns = ["basic_info.name", "basic_info.primary_domain",
           "basic_info.professional_network_url", "basic_info.employee_count_range",
           "locations.headquarters", "locations.country"]
result = subprocess.run(
    ["deepline", "run", "crustdata_v3_company_search",
     "--input", json.dumps(payload),
     "--fields", ",".join("companies." + c for c in columns),
     "--csv-out", "-"],
    capture_output=True, text=True, check=True
)
accounts = list(csv.DictReader(io.StringIO(result.stdout)))
# Each row: accounts[0]["basic_info.name"], accounts[0]["basic_info.primary_domain"], ...
```

**CrustData output → Clay field mapping:**

| CrustData row path                    | Clay formula field                                |
| ------------------------------------- | ------------------------------------------------- |
| `basic_info.name`                     | Name                                              |
| `basic_info.primary_domain`           | Domain                                            |
| `basic_info.professional_network_url` | LinkedIn URL                                      |
| `basic_info.employee_count_range`     | Size                                              |
| `locations.headquarters`              | Location                                          |
| `locations.country`                   | Country                                           |
| —                                     | Description, Primary Industry, Type (need Pass 2) |

### Pass 2 — Enrich missing fields (`prospeo_enrich_company`, optional)

Only needed if downstream passes reference `Description`, `Primary Industry`, or `Type`.

```python
# For each company from Pass 1 that lacks description/industry:
result = subprocess.run(
    ["deepline", "run", "prospeo_enrich_company",
     "--input", json.dumps({"website": domain}), "--json-out", "-"],
    capture_output=True, text=True, check=True
)
enriched = json.loads(result.stdout).get("company", {})
# Fields: description, employee_count, employee_range (= Size), industry, type
```

### Cost comparison

| Clay                                          | Deepline                                                                                             |
| --------------------------------------------- | ---------------------------------------------------------------------------------------------------- |
| Mixrank source — bundled in Clay subscription | CrustData company search + optional Prospeo enrichment; check live tool catalog for Deepline credits |
| Returns all fields in one step                | Two passes; Pass 2 optional if downstream uses only domain/name/linkedin                             |

### Key questions to ask the user before generating the script

1. What filters did the Clay Mixrank source use? (location, size, industry, tech stack) — visible in the Clay source config or ask the user
2. How many companies total? Use a narrow pilot or count-like query before paging.
3. Does the pipeline actually use `Description`, `Industry`, `Type`? If not, skip Pass 2.

---

## Company Name Normalization

### `normalize-company-name`

For LLM-quality normalization:

```ts
.withColumn('normalized_company', 'deeplineagent', { model: "openai/gpt-5.6-luna", prompt: "Normalize this company name: {{company_raw}}. Strip legal suffixes (Inc, LLC, Corp, Ltd, Holdings, Group). Return title case. Return ONLY the name, nothing else." })

```

For pure string normalization (cheaper):

```javascript
// $WORKDIR/normalize_company.js
const name = row['company_raw'] || '';
return name
  .replace(
    /\b(Inc\.?|LLC\.?|Corp\.?|Ltd\.?|Limited|Co\.?|Group|Holdings?)\b/gi,
    '',
  )
  .trim();
```

---

## Classification

### `chat-gpt-schema-mapper`

```ts
.withColumn('job_function', 'deeplineagent', { model: "openai/gpt-5.6-luna", prompt: "Classify this job title into a function label. Rules: all lowercase except BizOps/RevOps/GTM Ops, <4 words, describe the vertical. Return ONLY the label.\n\nJob title: {{job_title}}" })

```

No `jsonSchema` needed for a single-value string output.

---

## AI Columns — No Web

### `use-ai` (useCase: `use-ai`, no web tools)

```ts
// fast reasoning / generation
.withColumn('data_warehouse_formatted', 'deeplineagent', { model: "openai/gpt-5.6-luna", prompt: "<exact Clay prompt with {{field}} refs translated>" })


// larger structured output
.withColumn('strategic_summary', 'deeplineagent', { model: "anthropic/claude-sonnet-4.6", prompt: "<prompt>", jsonSchema: { type: "object", properties: { response: { type: "string" }, top_5_initiatives: { type: "string" }, top_3_sales_initiatives: { type: "string" } }, required: ["response"], additionalProperties: false } })

```

Reference structured output fields in downstream passes as `{{col_name.field}}`. If you need deeper nesting, flatten it first.

---

## AI Columns — Claygent (Web Research)

### `use-ai` (useCase: `claygent` or web browsing enabled)

**Always two passes.** Never combine broad research + generation in one model step — it is harder to debug and less stable than a split search/synthesis flow.

**Pass 1 — Research:**

```ts
.withColumn('company_research', 'deeplineagent', { model: "openai/gpt-5.6-luna", prompt: "Use the research context to summarize {{company_domain}} ({{company_name}}). Return JSON with summary, initiatives, and sources.", jsonSchema: { type: "object", properties: { summary: { type: "string" }, initiatives: { type: "string" }, sources: { type: "string" } }, required: ["summary", "initiatives"], additionalProperties: false } })

```

**Alternative research via exa_search** (deterministic, auditable):

```ts
.withColumn('exa_search', 'exa_search', { query: "{{company_name}} {{company_domain}} strategic initiatives GTM 2024 2025", num_results: 5, contents: { text: true, highlights: true } })

```

**Pass 2 — Generation (a later play column reading `{{company_research}}`):**

```ts
.withColumn('strategic_initiatives', 'deeplineagent', { model: "anthropic/claude-sonnet-4.6", prompt: "<Clay prompt translated>\n\nResearch context:\n{{company_research}}", jsonSchema: { type: "object", properties: { top_5_initiatives: { type: "string" }, top_3_sales_initiatives: { type: "string" }, top_3_go_to_market_initiatives: { type: "string" }, new_products: { type: "string" }, hypothesis_of_potential_challenges: { type: "string" } }, required: ["top_5_initiatives"], additionalProperties: false } })

```

---

## Octave Actions (proprietary — search for native equivalent first)

Octave `ca_*` agents are proprietary Clay integrations. Before defaulting to `deeplineagent`, check whether Deepline has added a native equivalent:

```bash
deepline search "person profile -> ICP qualification"
deepline search "prospect and messaging -> generated email sequence"
```

If search finds a native tool for the action, use it directly — it will be faster and more accurate than the `deeplineagent` fallbacks below. The patterns below are **fallbacks for when no native tool is available**.

### `octave-qualify-person`

```ts
.withColumn('qualify_person', 'deeplineagent', { model: "anthropic/claude-sonnet-4.6", prompt: "Score this prospect against our ICP (0-10 total). ICP: [paste ICP criteria]. Prospect — Title: {{title}}, Company: {{company_name}}, Domain: {{company_domain}}, LinkedIn: {{linkedin_url}}, Initiatives: {{strategic_initiatives}}.\n\nScoring dimensions: persona fit (0-4) + seniority (0-2) + hiring signals (0-2) + strategic fit (0-2).\nTier: A=8-10, B=5-7, C=0-4. Qualified if score>=6.\nReturn JSON.", jsonSchema: { type: "object", properties: { score: { type: "number" }, tier: { type: "string", enum: ["A", "B", "C"] }, qualified: { type: "boolean" }, rationale: { type: "string" }, disqualifiers: { type: "array", items: { type: "string" } } }, required: ["score", "tier", "qualified", "rationale"], additionalProperties: false } })

```

### `octave-enrich-person`

```ts
.withColumn('person_enriched', 'deeplineagent', { model: "anthropic/claude-sonnet-4.6", prompt: "Research this person using public sources. Name: {{first_name}} {{last_name}}, Title: {{title}}, Company: {{company_name}}, LinkedIn: {{linkedin_url}}. Return JSON with background, career_summary, and notable_achievements.", jsonSchema: { type: "object", properties: { background: { type: "string" }, career_summary: { type: "string" }, notable_achievements: { type: "string" } }, required: ["background", "career_summary"], additionalProperties: false } })

```

### `octave-run-sequence-runner` (email generation)

Two passes:

**Pass 1 — Gather signals (separate enrich call):**

```ts
.withColumn('sequence_signals', 'deeplineagent', { model: "openai/gpt-5.6-luna", prompt: "Summarize outbound signals for {{first_name}} ({{title}} at {{company_name}}). Use context: {{qualify_person}} / {{strategic_initiatives}} / {{tension_mapping}}. Return key talking points and pain hypotheses." })

```

**Pass 2 — Write email:**

```ts
.withColumn('email_sequence', 'deeplineagent', { model: "anthropic/claude-sonnet-4.6", prompt: "Write a cold email (subject + body, body <70 words). Recipient: {{first_name}}, {{title}}, {{company_name}}. Signals: {{sequence_signals}}. Tone: casual, direct. No buzzwords.", jsonSchema: { type: "object", properties: { subject: { type: "string" }, body: { type: "string" } }, required: ["subject", "body"], additionalProperties: false } })

```

---

## LinkedIn Posts

### `social-posts-get-post-activity-posts-and-shares`

Use a user-provided post export that the user is authorized to access and process. If one is unavailable, leave this step unresolved. Do not substitute a scraper or browser-session collection workflow.

---

## Scoring

### `score-your-data` (unconfigured — all input slots blank)

Replace with `run_javascript` keyword scorer. **Column names below are placeholders — replace with your actual Clay column aliases from the flatten pass.**

```javascript
// $WORKDIR/score_row.js
// Replace field names with your actual column aliases (from fields.xxx or top-level)
let score = 0;
const title = (row['job_title'] || '').toLowerCase(); // ← your title column
const signal1 = (row['hiring_signal'] || '').toLowerCase(); // ← your first signal column
const signal2 = row['tech_stack']; // ← your second signal column

// Adjust keywords and weights to your ICP scoring criteria
if (['vp', 'director', 'head of'].some((k) => title.includes(k))) score += 3;
if (signal1 && signal1.length > 10) score += 2;
if (signal2) score += 2;

const tier = score >= 7 ? 'A' : score >= 4 ? 'B' : 'C';
return { score, tier };
```

---

## CRM Read and Write

Clay ships 27 HubSpot actions plus a Salesforce package, and this file does not enumerate them. CRM read/write mappings, plus the two Clay-native table actions (`lookup-row-in-other-table`, `route-row`), live in [clay-api-surface.md](clay-api-surface.md#crm-read-and-write). Go there for any `actionKey` not mapped above.

---

## Campaign Activation

### `add-lead-to-campaign` (Smartlead)

Use `smartlead_push_to_campaign`. A single CLI push takes literal values for one
lead; check it with `--dry-run`, then run the same command
without it once the user approves the write:

```bash
deepline run smartlead_push_to_campaign --dry-run --input '{
  "campaign_id": "12345",
  "lead_list": [
    {
      "email": "ada@example.com",
      "first_name": "Ada",
      "last_name": "Lovelace",
      "company_name": "Example Co",
      "linkedin_profile": "https://www.linkedin.com/in/ada-lovelace-example"
    }
  ]
}'
```

Per-row pushes belong in a play column, which reads each row's values:

```ts
.withColumn('campaign_push', 'smartlead_push_to_campaign', { campaign_id: '<campaign_id>', lead_list: [{ email: '{{final_email}}', first_name: '{{first_name}}', last_name: '{{last_name}}', company_name: '{{company_name}}' }] })

```

### `add-lead-to-campaign` (Instantly)

Use `instantly_add_to_campaign`. As with Smartlead, check a single literal lead
with `--dry-run` first, run it for real only after the user approves the write,
and put per-row pushes in a play column:

```bash
deepline run instantly_add_to_campaign --dry-run --input '{
  "campaign_id": "8f3c2a1e-5b7d-4e9a-9c1f-2d6e8a4b7c30",
  "leads": [{"email": "ada@example.com", "first_name": "Ada", "last_name": "Lovelace", "company_name": "Example Co"}]
}'
```

### Other campaign platforms

| Platform                      | Tool                         | Notes                       |
| ----------------------------- | ---------------------------- | --------------------------- |
| HeyReach (LinkedIn sequences) | `heyreach_add_to_campaign`   | LinkedIn outreach sequences |
| Lemlist                       | `lemlist_add_to_campaign`    | Multi-channel sequences     |
| Smartlead                     | `smartlead_push_to_campaign` | Email sequences             |

---

## Field Reference Translation

Clay uses `{{f_0sy80p3xxx}}` field IDs in prompts and formula cells. Steps to translate:

1. Get field list from `GET /v3/tables/{TABLE_ID}` → `fields[].id` + `fields[].name`
2. Build the ID→name map: `f_0sy80p3xxx` → `snake_case(field.name)`
3. In recovered prompts, replace every `{{f_xxx}}` with `{{fields.snake_name}}` (post-flatten) or `{{alias}}` (if it's a prior pass output)
4. Fix Clay formula bugs sometimes present in rendered cell values:
   - Wrong field reference: `{{last_name}}` where `{{job_title}}` was intended — cross-check against the column name
   - Unresolved single-brace: `{field_name}` (Clay uses `{{double_braces}}` only) — add the second brace pair
5. Reference rules by output type:

| Source column type                   | Downstream reference                   |
| ------------------------------------ | -------------------------------------- |
| `run_javascript` returning a scalar  | `{{alias}}`                            |
| `run_javascript` returning an object | `{{alias.field_name}}`                 |
| `deeplineagent` without `jsonSchema` | `{{alias}}` (raw string)               |
| `deeplineagent` with `jsonSchema`    | `{{alias.field_name}}` for flat fields |
| Flattened Clay field                 | `{{fields.snake_name}}`                |

**Full path:** `{{alias.field}}`, `{{alias.field.nested}}`, and array indices like `{{alias.items[0].field}}` all resolve. A path that does not exist renders empty rather than erroring, so check a sample row when a value comes back blank.

---

## Column Alias Convention

Aliases **derive from the actual Clay column name**, not from a fixed list:

- Snake_case the Clay column name: "Work Email" → `work_email`, "Strategic Initiatives" → `strategic_initiatives`
- Strip leading/trailing spaces and special characters before snake_casing
- For multi-step patterns (email waterfall fallbacks), append a short functional suffix: `work_email_li` (LinkedIn fallback), `work_email_valid` (validation gate)

**Two reserved structural aliases** (always these names, regardless of Clay column names):

| Alias         | Purpose                                                       |
| ------------- | ------------------------------------------------------------- |
| `clay_record` | Raw bulk-fetch-records output loaded by the shell fetch pass  |
| `fields`      | Flattened clay_record subfields (run_javascript flatten pass) |

**All other aliases come from the Clay schema.** Look up `fields[].name` in `GET /v3/tables/{id}` and snake_case them. Do not invent aliases from any memorized list — if the Clay column is named "Tension Mapping", the alias is `tension_mapping`. If it's named "PVP Messages", it's `pvp_messages`.
deprecation-and-migrationaddyosmani/agent-skillsManages deprecation and migration. Use when removing old systems, APIs, or features. Use when migrating users from one implementation to another. Use when migrating a database schema in production, such as renaming or dropping a column without downtime (expand/contract). Use when deciding whether to maintain or sunset existing code.Stars: 119kUpdated Claude CodeCodexVerifiedInstall
performance-optimizationaddyosmani/agent-skillsOptimizes application performance across frontend, backend, queries, and databases. Use when performance requirements exist, when you suspect performance regressions, when Core Web Vitals or load times need improvement, when N+1 query patterns need fixing, or when profiling reveals bottlenecks.Stars: 119kUpdated Claude CodeCodexVerifiedInstall
agent-cost-reportthedotmack/claude-memBelievable agent cost report for any period, default the last 7 full days PT, not counting today. Measured tokens from Claude Code transcripts priced at OpenRouter list prices (ESTIMATED), measured provider spend when a sanctioned source exists, note-taker cost separate, Timing-style HTML/PDF plus report.json, line-items.csv, evidence.json.Stars: 118kUpdated Claude CodeCodexVerifiedInstall
agent-cost-reportthedotmack/claude-memBelievable agent cost report for any period, default the last 7 full days PT, not counting today. Measured tokens from Claude Code transcripts priced at OpenRouter list prices (ESTIMATED), measured provider spend when a sanctioned source exists, note-taker cost separate, Timing-style HTML/PDF plus report.json, line-items.csv, evidence.json.Stars: 118kUpdated Claude CodeCodexVerifiedInstall