本文へ移動
cccskills
無料GitHub で公開

pmstudio-comms

Generate templated stakeholder communications from project context. Use when someone asks to "write a go-live email", "create a status update", "onboard someone", "write an announcement", "steerco update", "incident summary", "change notification", or any stakeholder communication task. Reads project memory, stakeholder directory, PRD, and meeting notes to pre-fill real names, dates, and context. Produces ready-to-send Markdown.

インストール方法を見る

含まれるファイル(2)

  • SKILL.md4.9 KB
  • references/templates.md5.6 KB

SKILL.md(原文)

インストールする前に、エージェントに与えられる指示の中身を確認できます。

Stakeholder Comms — Templated Communications Generator

Purpose

Generates stakeholder communications pre-filled with project context. Not generic templates — these pull real stakeholder names, real dates, real scope from project artifacts.

Process

Step 1: Identify Template

From user request or by asking:

CommandTemplateWhen to Use
/stakeholder-comms go-liveGo-Live AnnouncementProduct/feature launching
/stakeholder-comms statusStatus UpdatePeriodic reporting (weekly/biweekly)
/stakeholder-comms onboard {name}Onboarding BriefNew team member joining
/stakeholder-comms changeChange NotificationPlanned change to live product
/stakeholder-comms incidentIncident SummaryPost-incident communication
/stakeholder-comms steercoSteerCo UpdateLeadership/governance meeting prep

If no subcommand, ask: "What type of communication do you need?"

Step 2: Read Context

All templates read:

  • CLAUDE.local.md — project overview, key decisions, stakeholders, recent changes
  • .sync-watch.json — project name

Template-specific reads:

TemplateAdditional Sources
Go-LivePRD (scope, success metrics), Meeting-Notes (launch decisions)
StatusPRD (milestones, timeline), Meeting-Notes (recent), Change-Log (recent entries)
OnboardStakeholder Directory, PRD (overview), Presentations (links to share)
ChangePRD (affected components), DR-Plan (rollback procedures if exists)
IncidentIRP (severity definitions if exists), Meeting-Notes (incident details)
SteerCoPRD (timeline, risks), Change-Log (period changes), all recent Meeting-Notes

Step 3: Ask Clarifying Questions

Only ask what can't be inferred from context:

  • Audience: Who receives this? (all stakeholders / specific group / leadership)
  • Tone: Formal (board) / professional (cross-team) / casual (own team)
  • Period: For status updates — what date range?
  • Specific content: For incident — what happened? For change — what's changing?

Step 4: Generate Communication

Output location: Comms/{Type}-{Date}-{Audience}.md

If Comms/ directory doesn't exist, create it.

Templates

See references/templates.md for the full template content for each type. Summary:

Go-Live Announcement

  • Subject line, TL;DR (2 sentences), What's Live (scope), Who It Affects, What To Do (action items per audience group), Known Limitations, Support Contacts, Timeline (what's next)

Status Update

  • Period, RAG Status (Red/Amber/Green with explanation), Key Accomplishments (bulleted), Decisions Made, Risks & Blockers (with owners), Upcoming Milestones (next 2 weeks), Action Items (table: item/owner/due), Help Needed

Onboarding Brief

  • Welcome, Product Overview (2-paragraph summary), Your Role, Key Contacts (table: name/role/email/when-to-contact), Access Setup Checklist, Key Documents (links), First-Week Actions, Glossary (10 most-used terms)

Change Notification

  • What's Changing, When, Why, Who's Affected, What You Need To Do, Rollback Plan, Support Contacts

Incident Summary

  • Severity, Timeline (bullet list: detected/responded/resolved), Impact, Root Cause, Resolution, Prevention (what we're doing so this doesn't happen again), Action Items

SteerCo Update

  • Executive Summary (3 bullets max), RAG Status with trend arrow, Decisions Needed (numbered, with context and recommendation), Budget & Timeline (table), Key Risks (top 3 with mitigation status), Accomplishments Since Last SteerCo, Next Period Focus

Step 5: Present and Refine

Show the generated communication. Ask:

  • "Ready to save, or would you like to adjust anything?"
  • For emails: "Want me to include a subject line suggestion?"

Critical Rules

  1. Use real data. Pull actual names, dates, and scope from project files. Never use placeholders like "[Name]" if the information exists in context.
  2. Appropriate detail level. SteerCo updates are concise (1 page). Onboarding briefs are comprehensive. Match the template.
  3. Action-oriented. Every communication should make clear what the reader should DO. No information-only dumps.
  4. Tone matching. Leadership comms use formal language. Team comms can be direct. Incident comms are factual and calm.
  5. Don't fabricate. If you don't have incident details, ask. Don't guess root causes or timelines.

レビュー

まだレビューはありません。使ってみた感想をお寄せください。

同じリポジトリのスキル

概要と使いどころ

Train and optimize AI agents using Microsoft's Agent Lightning framework with reinforcement learning. Use when setting up agent training, instrumenting agents with tracing, configuring LightningStore, implementing reward functions, or optimizing prompts with RL/APO algorithms.

日本語の概要は準備中です。原文の説明を表示しています。

coco-research/coco5512026年10月11日 更新

Post-run self-evaluation system that scores agent output on correctness, clarity, actionability, and conciseness. Use after /team runs, skill executions, or when explicitly asked to evaluate output quality.

日本語の概要は準備中です。原文の説明を表示しています。

coco-research/coco5512026年10月11日 更新

Create AI marketing videos for ads, promos, product launches, and brand content. Models: Veo, Seedance, Wan, FLUX for visuals, Kokoro for voiceover. Types: product demos, testimonials, explainers, social ads, brand videos. Use for: Facebook ads, YouTube ads, product launches, brand awareness. Triggers: marketing video, ad video, promo video, commercial, brand video, product video, explainer video, ad creative, video ad, facebook ad video, youtube ad, instagram ad, tiktok ad, promotional video, launch video

日本語の概要は準備中です。原文の説明を表示しています。

coco-research/coco5512026年10月11日 更新

Use when building AI features into a product: LLM integration, RAG pipelines, guardrails, streaming, AI UX, prompt engineering, or AI cost control. Treats prompts as code and validates every model output.

日本語の概要は準備中です。原文の説明を表示しています。

coco-research/coco5512026年10月11日 更新

Your AI research and engineering brain trust. 59 named personas across 8 cells covering frontier labs, applied product, model architecture, reasoning/RL/agents, alignment and interpretability, theory and science of DL, multimodal and…

日本語の概要は準備中です。原文の説明を表示しています。

coco-research/coco5512026年10月11日 更新

Use when designing a new REST or GraphQL API, reviewing an API spec before implementation, setting team API standards, or migrating REST to GraphQL. Covers resources, HTTP semantics, pagination, error handling, and pitfalls.

日本語の概要は準備中です。原文の説明を表示しています。

coco-research/coco5512026年10月11日 更新

coco-research のスキルをすべて見る

このスキルの問題を報告する