deepline-plays-improve

Solid

Use this skill when a Deepline Play needs human review, evaluation against golden labels, or bounded iterative improvement. Triggers on “put this run in a Sheet,” “review these results,” “read my feedback,” “compare this run to the golden set,” “evaluate this play,” “optimize this play,” or “add the revised run to the same spreadsheet.” It covers fresh Google Sheets feedback reads, durable Play Dataset exports, golden-set discipline, and safe iteration through `deepline tools execute`. Skip ordinary CSV export with no review or evaluation loop.

AI & Automation 47 stars 11 forks Updated today MIT

Install

View on GitHub

Quality Score: 83/100

Stars 20%
56
Recency 20%
100
Frontmatter 20%
70
Documentation 15%
100
Issue Health 10%
50
License 10%
100
Description 5%
100

Skill Content

# Deepline Plays Improve ## Quick Start ```bash npm install -g deepline # Fallback for secure sandboxes: mkdir -p "$HOME/.local" && npm config set prefix "$HOME/.local" && export PATH="$HOME/.local/bin:$PATH" && npm install -g deepline --registry https://code.deepline.com/api/v2/npm/ deepline auth register --wait auto deepline auth wait --timeout 120 # completes Cowork/browser approval; no-op if already connected deepline auth status deepline -h ``` Improve a Play through one of three related modes. Google Sheets is the review and labeling surface; completed Play datasets remain the execution record. | User intent | Mode | Default behavior | | ------------------------------------- | -------- | --------------------------------------------------------- | | Review one run or address comments | Review | One feedback pass, then a revised run | | Measure a run against expected labels | Evaluate | Score once against a fixed golden-data version | | Iteratively improve a metric | Optimize | Bounded candidate loop with development and holdout cases | Do not silently turn review into optimization. Optimization makes repeated provider calls and changes Play behavior, so it needs an explicit metric, budget, and stopping rule. The live CLI is the source of truth. Confirm the tools before first use: ```bash test -n "${WORKDIR:-}" || { echo "Set WORKDIR to a writable t...

Details

Author
getaero-io
Repository
getaero-io/gtm-eng-skills
Created
5 months ago
Last Updated
today
Language
TypeScript
License
MIT

Integrates with

Similar Skills

Semantically similar based on skill content — not just same category

AI & Automation Solid

deepline-plays-review

Use this skill when a human needs to review a Deepline Play result and hand feedback, labels, or approval back to the agent for revision, evaluation, comparison, or bounded iteration. Triggers on “put this run in a Sheet,” “review these results,” “read my feedback,” “make this a standing rule,” “never regress on this case,” “compare these runs,” “keep improving,” or “add the revised run to the same spreadsheet.” Skip ordinary Play authoring or execution with no human review loop, generic spreadsheet work, and one-off CSV export.

47 Updated today
getaero-io
AI & Automation Solid

deepline-plays

Use this skill for any GTM Engineering work in Deepline — prospecting, list-building, account research, contact enrichment, email/phone/LinkedIn lookup, ICP qualification, lead scoring, outreach copy, CSV-driven row work, and verifying enriched data. Triggers on phrases like 'enrich this CSV', 'find contacts at these companies', 'build a TAM list', 'waterfall emails for these leads', 'detect job changes', 'is this data accurate', 'write a sequence for', and on any request that mentions Deepline, plays, or named GTM providers (Crustdata, Hunter, Dropleads, etc.). Use this even when the user does not explicitly say 'GTM' — most CSV-with-leads tasks and most provider-driven enrichment tasks are GTM tasks. SKIP only when the request is a Clay table extraction (use clay-to-deepline) or has no Deepline / outbound / data-enrichment dimension at all.

47 Updated today
getaero-io
AI & Automation Solid

deepline-gtm

GTM prospecting, enrichment, research, outreach, scoring, CSVs, and Deepline plays. Discovery: deepline-pre-research. Providers: adyntel, ai_ark, allegrow, apify, attention, attio, aviato, bettercontact, bloomberry, bluesky, bounceban, browserbase, builtwith, clickhouse, cloudflare, contactout, crustdata, crustdata-v2, crustdata-v3, customer_db, dataforseo, datagma, deepline_native, deeplineagent, discolike, dropleads, emailbison, emailguard, enformion, exa, findymail, firecrawl, firmable, forager, fullenrich, generic_http, gong, google_ads_audiences, google_workspace, hackernews, heyreach, hubspot, hunter, icypeas, instantly, intercom, ipqs, leadmagic, lemlist, limadata, linkedin_ads_audiences, linkedin_scraper, lusha, meta_audiences, openmart, opensosdata, openwebninja, parallel, peopledatalabs, podscan, predictleads, prospeo, rocketreach, salesforce, salesforge, scrapecreators, sentrion, serper, slack, smartlead, snowflake, sumble, theirstack, trestle, twitterapi, upcell, wiza, wizleads, zerobounce.

47 Updated today
getaero-io