← ClaudeAtlas

antwork-auditlisted

Use when the user wants a full audit, health check, or scorecard of their social-media presence managed through Antwork — engagement performance, voice consistency, posting cadence, content quality, and platform coverage. Trigger on "audit my socials", "how's my LinkedIn/X doing", "social health check", "review my Antwork account", "what should I fix", "grade my content".
polynode-co/antwork-skills · ★ 1 · AI & Automation · score 67
Install: claude install-skill polynode-co/antwork-skills
# Antwork Social Presence Audit A comprehensive, data-driven audit of everything the user ships through Antwork. It mirrors a 5-dimension marketing audit: five specialized agents run in parallel, each owning one dimension, then you synthesize a single scored report. ## 1. Establish scope first Before spawning anything, gather the lay of the land so the agents (and the report) reference real accounts and real numbers: - `list_workspaces` — confirm the default workspace (the user's plan and quota are in its `account` field). If there is more than one and no default, ask which to audit (or `set_default_workspace`). - `list_social_accounts` — the connected platforms, handles, token health, and each account's latest `followers` (count and 30-day change). This is the audit's universe. - `get_post_context` — its `workspace` block: brand identity, timezone, preferred posting times. If no accounts are connected, stop: there's nothing to audit yet. Call `get_connection_urls` so the user can connect one. ## 2. Spawn the five dimension agents in parallel Launch all five in a single batch so they run concurrently. Each is read-only and returns a structured findings block with a 0–100 sub-score and its top 3 fixes. Pass each the workspace and the account list you gathered. **No subagents where you are running** (claude.ai, or any host without a subagent tool)? Run the five dimensions yourself, one after another, following each agent's instructions file (`agents/<name>.md` in this s