Codex GTM Skill Pack

Turn GTM knowledge into a ranked action plan.

Install the workflow, answer a focused intake, and let Codex turn your market knowledge into signal-backed account ranking, top 5 lead groups, an outreach playbook, and an approval-ready Apollo handoff.

2026-04-30.2Pack version
v1.2.0Codex-only workflow contract
10 skillsInstalled into Codex
9 outputsExample run outputs

Setup Flow

1

Unzip

Open the folder and run setup from the root directory.

2

Install

Global install copies skills and support assets into Codex.

3

Restart

Restart Codex so the new skills are discoverable.

4

Smoke Test

Run the scorer against the included example data. Top row should be Verglas Cybersecurity at 83.50.

5

Trigger

Answer the GTM Q&A first. Scoring starts only after the workflow understands your market, signals, and approval rules.

cd chatgpt-gtm-codex-skill-pack ./install.sh # optional tenant fork cp -R profiles/_template profiles/acme ./install.sh --profile acme # smoke test python3 skills/chatgpt-account-prioritization/scripts/score_accounts.py examples/worked-run/accounts.csv

Trigger Prompt

Use chatgpt-gtm-orchestrator. Start with the interactive GTM Q&A before any analysis. Ingest what I know, identify gaps, align on supplied list or cold audience, define target leads, signals, triggers, Apollo live search and AI Research flow, preloaded skill map, approval gates, then produce the first run plan.

The workflow turns what you know into a usable plan: gaps, approved sources, live-search path, signal model, ranked accounts, top 5 lead groups, outreach playbook, Apollo action queue, and QA.

What You Get

Intake summary Gap analysis Source matrix Live search plan Apollo search plan Ranked accounts Top 5 lead groups GTME playbook Sequence fit review Apollo action queue QA report
Demo output files: accounts.csv, accounts_ranked.csv, top_leads.md, gtme_playbook.md, skill_task_map.md, sequence_review.md, apollo_action_queue.csv, qa_report.md, intake_summary.json

Workflow Map

Intake

Q&A captures GTM knowledge, supplied or cold audience, targets, signals, tools, skills, and approvals.

Analyze

Connector audit, source mapping, gap analysis, and the scoring plan.

Search

Apollo live search, AI Research, and connected tools only when approved.

Score

Deterministic account scoring, lead-group selection, top 5, and rejection reasons.

Handoff

Playbook, sequence fit, Apollo action queue, meeting dossier when needed, and QA verdict.

Approval Boundary

Allowed before approval

  • Read-only analysis
  • Draft outreach plan
  • Draft Apollo action queue
  • Example scoring run

Held until approved

  • Enrichment credit use
  • Apollo or CRM writes
  • Sequence membership changes
  • Message sends, uploads, external shares

Built-In Safeguards

  • Scoring weights can be calibrated before rankings are used for decisions.
  • Example data is synthetic and provided for setup validation only.
  • Search and scoring stay tied to approved sources, signals, and approval rules.
  • Writes and sends require explicit action-level approval.