Legacy content framework. Use only when the user explicitly invokes $audience-content-os. Do not auto-trigger for ordinary content, LinkedIn, newsletter, hook, headline, audit, or source-transformation requests.
Turn an approved B2B market and offer hypothesis into evidence-backed account-fit rules, buyer-role rules, signal criteria, list-source choices, exact pull filters, QA rules, sample-review plans, and campaign-ready sourcing specifications. Use when Codex needs to research who should enter an outbound list, define a defensible reason for contact, compare sourcing paths, prepare a human-reviewed list brief, or translate already-approved rules into a campaign creation brief/config before any scraping, campaign mutation, launch, or send. Do not use for copy-only requests or to execute scraping, enrichment, campaign mutation, launch, or sending; route those actions to their dedicated skills after the required approval.
Audit an existing B2B offer from the perspective of a cold buyer with no transferred trust. Use when reviewing a website, proposal, call transcript, case study, offer note, pricing or terms, or user-supplied answers to determine what is clear, supported, coherent, and safe to test; expose contradictions, claim-to-proof gaps, risk, effort, time-to-value, and first-commitment weaknesses without claiming market validation or likely conversion. Do not use to invent, redesign, or build an offer; route an explicitly requested post-audit rebuild to offer-creation.
Collect Google Maps business listings by category and location through the RapidAPI Maps Data API. Use when you need a deduplicated CSV of local companies with names, websites, phones, addresses, ratings, coordinates, and categories before qualification or contact research.
Build and tune a reusable prompt for deciding whether companies match an ideal customer profile. Use before applying AI qualification to a large company list. Test ten companies per round, incorporate user corrections, and stop only after two consecutive rounds require no corrections.
META skill — build the largest possible qualified lead list for any request, end to end. Orchestrates discovery (Prospeo + GetLeads + Blitz + Google Maps + lookalike snowball from /list-expander), AI qualification with live-website verify, uncapped contact pull (GetLeads → Blitz → Prospeo), and provider emails (GetLeads/Blitz/Prospeo) validated with MillionVerifier before sending. Use for any "build me a list", "get me leads for X", "find everyone who...", client list request, or campaign list build. Starts with a 7-line intake brief; outputs to Postgres, Google Sheet, and/or CSV (asks which).
Expand a hard-to-build niche list from ~10 known-good seed companies into a full qualified TAM. Seed fingerprinting (how do good-fit companies ACTUALLY show up in the company database's filters) → lookalike generation (Prospeo company_lookalike, Exa findSimilar, Parallel.ai entity search) → filter mining with a precision/volume scorecard → wide pull with auto-sharding → cheap-AI qualification → live-website verification → client transparency report. Use when a vertical list feels "too small" (medical groups, hedge funds, Medicare brokerages), when known good-fit companies don't show up in keyword searches, or when someone says "expand this list", "the TAM should be bigger", "find more companies like these".
Create, improve, reposition, and evaluate business offers using live market research plus structured offer design. Use when the task is to: (1) turn a topic, niche, or problem into an offer, (2) improve an existing offer, (3) adapt an offer to a new market or segment, (4) choose pricing, guarantees, tiers, or qualification rules, or (5) analyze whether an offer is strong enough to sell.
Audit first-party B2B outbound cohorts and calculate channel-correct funnel conversion, leakage, fully loaded CAC, first-year gross profit, gross-profit LTV:CAC, CAC payback, break-even customers, required customer rate, and marginal expansion scenarios. Use when Codex must analyze cold-email, LinkedIn, or named-account outbound results; normalize CRM or campaign exports; compare cohorts; plan a pilot without invented benchmarks; test whether acquired customers can support acquisition cost; or produce an evidence-led funnel and economics report. Preserve missing data as unknown, keep actuals separate from projections, and never infer causation from a rate.
Turn a verified TikTok slideshow reference into editable 9:16 HTML slides and local PNG files. Preserve the reference's structure, use only supported claims, render in the browser, and complete a manual visual and posting check.
Turn one complete source, draft, or verified idea into a source-faithful X post. Choose the native format, surface the sharpest true hook, preserve every material fact, and return a finished post without invented experience or claims.
Plain text files in a repository that tell a coding agent how the project works: commands to run, conventions to follow and things to avoid. CLAUDE.md, AGENTS.md, cursor rules and skills are the common kinds.
CLAUDE.md or AGENTS.md?
CLAUDE.md is read by Claude Code. AGENTS.md is an open format that Codex, Cursor and other agents read. Many projects keep one and point the other at it.
What is a skill?
A folder with a SKILL.md that describes one capability, such as filling PDFs or reviewing code. The agent loads it only when the task calls for it.
Can I search my own team's files too?
Your agents already can, over MCP, limited to the files you're allowed to read. Searching them from this page is coming.