generate-account-plan — Claude Code command: overview and setup
The generate-account-plan Claude Code command builds a strategic plan for a company account. It reviews account, contact, and activity data, with web search as a supplement when requested or when data is sparse.
Apache-2.0 ★ 36,654 updated 2026-10-24
/plugin marketplace add anthropics/knowledge-work-plugins
/plugin install common-room@knowledge-work-pluginsWhat generate-account-plan does
The command turns a company name or domain into an account plan with an executive summary, company overview, stakeholder map, activity analysis, opportunities, risks, and action items. It ranks the top five contacts by member score and assigns each a stakeholder role. Activity analysis covers the last 90 days, with up to 50 events.
Web search runs when Common Room data is sparse or the user asks for it; it looks for company news from the last 30 days. When the account has little data, the command keeps only sections supported by evidence and shortens the plan. If references/my-company-context.md is available, it uses that context to tailor the summary and action items.
What it helps with
- build an account plan for a named company
- identify key contacts and classify their roles
- summarize account engagement over time
- turn account signals into opportunities, risks, and next steps
Example prompts for the agent
- Create an account plan for Acme using the domain
acme.com. - Review Acme’s contacts and activity from the last 90 days.
- Prepare an account plan for Acme and add recent company news.
How to install generate-account-plan
1Plugin marketplace
/plugin marketplace add anthropics/knowledge-work-plugins/plugin install common-room@knowledge-work-plugins2Manually: put the files in place
.claude/commands/generate-account-plan.mdCall it with /generate-account-plan. For all projects use ~/.claude/commands/
Manually: put the files in place
~/.codex/prompts/generate-account-plan.mdCall it with /prompts:generate-account-plan. Codex marks this format deprecated in favour of skills
Manually: put the files in place
.github/prompts/generate-account-plan.prompt.mdManually: put the files in place
.gemini/commands/generate-account-plan.tomlGemini stores commands as TOML: the prompt text goes into the prompt field
Questions about generate-account-plan
How do I install generate-account-plan in Claude Code?
Add the marketplace with /plugin marketplace add anthropics/knowledge-work-plugins, then install the plugin with /plugin install common-room@knowledge-work-plugins.
What does generate-account-plan do?
It reviews a company account, its contacts, and recent activity, then produces a structured plan with opportunities, risks, and action items. When evidence is sparse, it includes only sections supported by available data.
When does generate-account-plan use web search?
It searches for company news from the last 30 days when Common Room data is sparse or the user requests a search. If references/my-company-context.md is available, the command uses it to tailor the plan.
Source text
generate-account-plan.md · Apache-2.0
--- description: Generate a comprehensive strategic account plan argument-hint: <company name or domain> --- Generate a comprehensive account plan for "$ARGUMENTS". ## Process 1. **Full account research** — Follow the account-research skill to produce a complete account overview. Pull all available first-party (product), second-party (community), and third-party (intent) signals, CRM data, scores, and RoomieAI research. 2. **Stakeholder mapping** — Fetch the top 5 contacts at this company sorted by member score descending. Classify each into: Champion, Economic Buyer, Influencer, End User, or Unknown. Use Spark persona data if available; if Spark is unavailable, infer from engagement patterns and activity recency. If no activity data exists either, classify as Unknown. 3. **Engagement analysis** — Pull all organization activity for the last 90 days (up to 50 activities). Identify trends: is engagement growing, stable, or declining? Which contacts are most active? What channels are they engaging through? 4. **Web search (supplementary)** — If CR data is rich, skip. If data is thin or the user requests it, search for recent company news (last 30 days): funding, acquisitions, product launches, leadership changes, competitive moves. 5. **Synthesis** — Combine all data into a structured account plan. When the user's company context is available (see `references/my-company-context.md`), tailor the executive summary, opportunities, and action items to the user's product and ICP. ## Output Format ``` ## Account Plan: [Company Name] ### Executive Summary [3-4 sentences: relationship status, key opportunity, primary risk, recommended priority] ### Account Overview | Field | Value | |-------|-------| | Industry | ... | | Size | ... | | Domain | ... | | CRM Owner | ... | | Opp Stage | ... | | ARR | ... | | Scores | ... | ### Stakeholder Map **Champions** - [Name] — [Title] — [Key signals, last activity date] **Economic Buyers** - [Name] — [Title] — [Key signals] **Influencers** - [Name] — [Title] — [Key signals] **End Users** - [Name] — [Title] — [Key signals] ### Engagement Analysis [Trend summary: growing/stable/declining, most active contacts, top channels, comparison to 90 days prior if data available] ### Recent News [If web search was run] [Web search findings with sources and dates] ### Opportunities 1. [Signal-backed opportunity with specific next step] 2. ... ### Risks 1. [Signal-backed risk with mitigation] 2. ... ### Prioritized Action Items 1. [Specific action] — [Owner suggestion] — [Timeline] 2. ... 3. ... ``` Ground every insight in actual data. Flag explicitly when data is thin or unavailable. **If Common Room returns sparse data for this account**, produce an abbreviated plan that covers only the sections with real data. Do not generate a full account plan from minimal input — a short honest plan with gaps clearly noted is far more useful than a comprehensive-looking plan built on fabricated details. Omit sections (Stakeholder Map, Engagement Analysis, Opportunities, Risks) entirely if there is no data to support them.
© Common Room · Apache-2.0 · anthropics/knowledge-work-plugins