From Superposition
Control candidate outreach and follow-up emails in Superposition — draft/send a follow-up, chase a non-responder, send a polite decline, hold or cancel a queued send, or stop the agent because a human is taking over the thread. Use for "draft a follow-up about availability", "reach out and ask if he wants to reschedule", "send an apologetic decline", "don't send a follow-up", "I'll handle my own email", or "if they've had three emails with no reply, archive them".
How this skill is triggered — by the user, by Claude, or both
Slash command
/superposition:manage-outreachThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
Operators constantly steer candidate email — send, hold, stop, or hand off. Email **sends** happen
Operators constantly steer candidate email — send, hold, stop, or hand off. Email sends happen
through the Master Agent (invoke_agent), which owns the delivery guardrails; your job is to read the
thread, get the intent exactly right, and avoid sending the wrong thing.
get_candidate_email_thread (full history, not just the latest) +
get_candidates (current stage, source.category). Scan for whether the thing being asked was
already answered — never re-ask an answered question.list_delayed_emails to see pending sends before you add or
cancel one.invoke_agent, threading sessionId:
"For candidate
<id>on job<id>: <draft and send a follow-up asking X / send a polite closure / cancel the queued send / do not email — human is taking over>. Keep it warm and concise."
source.category: "INBOUND") can't be cold-emailed — the agent can only
reply to an existing thread; first-stage outreach is system-driven. If there's no thread, say so.Queued a follow-up to [Mei Tan](https://app.superposition.ai/c/def-456) asking about availability (sends in ~1 min — say "cancel" to stop it).
npx claudepluginhub superpositionhq/ai-plugin --plugin superpositionProvides behavioral guidelines to reduce common LLM coding mistakes, focusing on simplicity, surgical changes, assumption surfacing, and verifiable success criteria.
Searches, retrieves, and installs Agent Skills from prompts.chat registry using MCP tools like search_skills and get_skill. Activates for finding skills, browsing catalogs, or extending Claude.
Creates, edits, and optimizes skills for Claude Code, including drafting, evaluating with test prompts, iterating on performance, and improving skill descriptions for better triggering accuracy.