prioritize-accounts

作者: openai

根據帳戶排名優先處理準備好回覆的管道,暫停進行中的行動,選出最佳可聯繫的聯絡人,並產出基於連接器的…

npx skills add https://github.com/openai/role-specific-plugins --skill prioritize-accounts

Prioritize Accounts

Prepare a sales person with a small, evidence-grounded account action view: which accounts deserve attention now, why they rose, who to engage, and what safe next step to take.

Common Skill Instructions

MANDATORY: If not already in context, read and adhere closely to plugins/sales/skills/index/SKILL.md## Cross-Skill Best Practices.

Account Ranking

Use this priority order when the account set, ranking basis, or motion is ambiguous.

  1. Explicit user request, named account list, owner scope, territory, ranking basis, or stated preference
  2. CRM-backed open pipeline or user-provided CRM-equivalent account truth
  3. Accounts with a source-grounded why-now signal, clear net_new or expansion branch, and reachable primary contact
  4. Accounts with meaningful deal value, timing, momentum, risk, or actionability
  5. Accounts without active motion, duplicate outreach risk, ownership ambiguity, or another hard suppressor
  6. Accounts that fit the rep's stated capacity window

Default broad requests to open pipeline deals, deal value / ACV adjusted by timing, momentum, risk, and actionability, with a mixed motion goal. Do not create the candidate universe from enrichment, public context, calendar, or messages alone.

Key Dependency Categories

It is critical to use all of these categories if you can, if any relevant plugins or connectors are available, to ensure high quality responses. If the user declines, or connectors aren't available, state that limitation in your final answer. You can also use other relevant connectors as needed.

  • [Blocking] ~~CRM for account truth, ownership, customer status, opportunities, stage, amount, forecast posture, contacts, and recent activity. It blocks unless an authoritative account universe and account truth are already grounded in context.
  • ~~Calendar for active-motion suppression and timing
  • ~~Meeting Transcripts for objections, buying process, and stakeholder continuity
  • ~~Email for recent engagement and promised next steps
  • ~~Internal Messaging for duplicate motion, blockers, and owners
  • ~~Knowledge & Files for ICP guidance, target lists, territory rules, account plans, sales plays, and suppression conventions
  • ~~Sales Intelligence for fit, contactability, tech stack, funding, hiring, and source-grounded why-now context after CRM anchors the set

Overall Rules

  • Always cite sources using hyperlinks when useful links are available.
  • Sources must list every material dependency category, its status, and whether it affected the recommendation.
  • Keep the account as the primary unit of work and prefer fewer executable rows over a noisy list.
  • Put each account in exactly one of Suggested Focus, Monitor, or Suppress Or Block. When an account has both actionable work and a suppressor such as duplicate-outreach risk, keep it in Suggested Focus and describe the constrained action there; use Suppress Or Block only when no rep action should be taken now.
  • CRM owns the account universe and account truth. Other sources can enrich, suppress, or re-rank after the universe is anchored.
  • Calendar, meeting notes, messages, documents, and Sales Intelligence may change timing, suppression, contact choice, confidence, or why-now context after CRM anchors the account set; they must not create or overwrite the account universe.
  • Never fabricate contacts, urgency, opportunity posture, why-now signals, ownership, or account facts.
  • Do not execute outreach, create records, update CRM, or send messages in this workflow unless the user explicitly requests the separate approved action.
  • Use references/suppression-and-fallbacks.md when a row may be suppressed, blocked, monitored, or recovered through a fallback lane.

Next Step Options

After the first output, offer the most relevant follow-up from the options below. Offer one clear transition, not a menu. Suggest ONLY these unless you are very confident another option is more useful:

  • Take the next logical action for one selected account, such as drafting a message, preparing an internal ask, or unblocking a dependency.
  • Draft CRM-ready updates for review to address missing, stale, or contradictory fields that affected the ranking.
  • Research one selected account more deeply when the next action depends on missing context.
  • Refine the ranking, suppression rules, or motion goal using the user's guidance.
  • Check whether a matching weekly automation already reruns this prioritize-accounts skill for the same account set; if none exists, offer to create one that gives the seller a weekly account worklist with why-now context and safe next actions.

Next steps to avoid:

  • Blanket outreach, bulk CRM writes, or taking action across the whole ranked list without explicit user direction.

Automation Offer Guard

For substantive ranking outputs, a weekly account-priority brief is the preferred automation offer when the user has a reusable account set, owner scope, territory, pipeline view, target list, or named watchlist. Frame the value in sales language: deciding where to spend the week, which accounts moved into focus, which should be monitored or suppressed, and what next action is safe to take without creating duplicate outreach or CRM churn.

The automation must be a scheduled rerun of this skill, not a separate custom digest. When creating or describing the automation, make the prompt call this skill directly and preserve the same account scope and ranking intent:

Use the Sales `prioritize-accounts` skill.
Rerun it weekly for the same seller account set: [owner scope, territory, pipeline view, target list, or watchlist].
Return the standard Account Action View when source links are available; otherwise return the Chat Fallback output.

account_set: "[same resolved account set]"
ranking_basis: "[same ranking basis or motion goal]"
motion_goal: "[net_new, expansion, mixed, or user-specified]"
suppression_rules: "[same known suppression rules when relevant]"

The recurring output should follow this skill's account-priority contract: Suggested Focus, Monitor, Suppress Or Block, Evidence Gaps, and Source & Run Details. Keep it read-only; it may recommend account actions, messages to draft, or CRM corrections for review, but must not execute outreach, create records, update CRM, or send messages unless the user separately asks and approves.

Before offering the weekly account-priority brief, check whether the user already has a matching local automation installed. Inspect local automation records under $CODEX_HOME/automations/*/automation.toml, or ~/.codex/automations/*/automation.toml when CODEX_HOME is unset, and match by name, prompt, skill name, cadence, account set, owner, territory, pipeline view, ranking basis, motion goal, or other stable scope details. Treat active and paused matches as already installed.

  • If a matching automation exists, do not suggest creating another one. Continue with the next most relevant non-automation follow-up.
  • If no matching automation exists, end with one clear offer to check/create a weekly rerun of prioritize-accounts for the same scope. Describe the recurring output as a seller-ready weekly worklist of accounts to focus, accounts to monitor, accounts to suppress, why each matters, and the recommended next move. Do not create or update the automation until the user explicitly agrees.
  • If the automation surface is unavailable, do not mention tool details; offer to help set up a recurring account-priority brief when automations are available.

Modes

1. Account Action View

  • This is the required default mode whenever stable connected-source or source-of-truth links are available.

  • Do not choose Chat Fallback merely because the candidate set is small, the result can fit in chat, or some links need to be constructed from a trusted connector-returned org or instance base URL.

  • Use for substantive ranking runs when CRM, another searchable connected source, or stable source-of-truth links can open useful row context.

  • Build Work now, Watch, Paused, Evidence Gaps, and Source & Run Details. Map Suggested Focus to Work now, Monitor to Watch, and Suppress Or Block to Paused so the pane and any chat readout stay semantically identical.

  • Use the bundled template exactly:

    assets/account-priority-pane.template.html

  • The template is standalone UTF-8 HTML with inline CSS and browser JavaScript only. It has no external assets, no script imports, no npm dependencies, no Node dependency, and no helper scripts.

  • Do not inline, rewrite, restyle, summarize, or recompose the template. The only allowed HTML mutation is replacing its exact data placeholder with the escaped payload described below.

View Generation Rules

  • Generate the view whenever stable connected-source or source-of-truth links are available. A pasted unlinked list stays in chat unless the user explicitly asks for an offline view.
  • Generate one local file in an OS-appropriate temporary directory:
    • macOS/Linux: ${TMPDIR:-/tmp}/prioritize-accounts/index.html
    • Windows: %TEMP%\prioritize-accounts\index.html
  • Build the account action package using Work now, Watch, Paused, Evidence Gaps, and Source & Run Details.
  • Serialize the three account groups under workNow, watch, and paused respectively.
  • Give every account row the seller-facing fields required by the payload schema: rank, account, motion, stage, value, whyItMatters, nextAction, owner, dueDate, confidence, and status. Include accountUrl, primaryContact, and evidence when grounded and available.
  • Keep row copy compact enough to scan: whyItMatters should explain the timing, value, momentum, or risk in one or two short sentences, and nextAction should name one immediate safe action, watch trigger, or unblocking step.
  • Never invent display fields. Use Value unavailable, Unassigned, or No date set when the connected source does not provide value, owner, or due-date truth.
  • Serialize the same package as JSON that conforms exactly to references/template-payload.schema.json. Treat the schema's required root keys and account-row fields as the canonical payload contract.
  • Escape the JSON for safe embedding in a script tag by replacing every < with \u003c.
  • Replace the exact template placeholder __PRIORITIZE_ACCOUNTS_DATA_JSON__ with the escaped JSON.
  • Write the resulting HTML as UTF-8 to the local output path.
  • Sanity check that the generated file contains no unresolved placeholder.
  • Return the actual clickable view link first in the final response, then a short summary and proposed next step. Do not hardcode macOS-only /private/tmp in Windows-facing output. URL-encode spaces if emitting a file:// URL.
  • If local file links do not open in the client, start a tiny local HTTP server yourself only when necessary and return the localhost URL.
  • Do not mention templates, JSON, placeholders, local files, or HTML mechanics in ordinary user-facing copy.

2. Chat Fallback

  • Use when no useful source search/linking is available, local view creation is impossible, the user asks for plain text, or a parent flow requires chat-only output.
  • State why the account action view was skipped when the missing source/link condition caused the fallback.

Output Format

# Prioritize Accounts

## Suggested Focus

| Account | Why Now | Primary Contact | Secondary Contact | Opp Recommendation | Suggested Next Step | Sequence Angle | Confidence | Status |
| --- | --- | --- | --- | --- | --- | --- | --- | --- |
| [Account + source link] | [Grounded timing reason] | [Name + role] | [Name + role or None] | [Practical recommendation] | [Immediate rep action] | [Compact outreach hook] | [High/Medium/Low] | [ready_now/partial] |

## Monitor

| Account | Reason To Monitor | Trigger To Revisit | Status |
| --- | --- | --- | --- |

## Suppress Or Block

| Account | Suppressor Or Blocker | Evidence | Status |
| --- | --- | --- | --- |

## Sources

- [Sources that were included in the analysis]
- [Sources that were not included, why, and potential next steps to resolve]

## Source & Run Details

- **Account set:** [Resolved set]
- **Ranking basis:** [Visible rule]
- **Source of truth:** [CRM or CRM-equivalent source]
- **Motion goal:** [net_new / expansion / mixed]
- **Sources checked:** [Sources]
- **Assumptions:** [Assumption-based only when applicable]

---

{Follow the instructions and output format/conditions in [Limitations and Improvements](../index/SKILL.md#limitations-and-improvements)}

{Follow the instructions and output format/conditions in [Next Steps](../index/SKILL.md#4-next-steps)}

來自 openai 的更多技能

user-context
openai
載入或管理資料分析插件的持久性來源路由偏好設定、入門邏輯、設定進度以及語義層註冊表。
official
notion-research-documentation
openai
研究Notion內容,並將其整合為結構化的簡報、報告或附有引用的比較分析。透過精準查詢搜尋並擷取Notion頁面,然後按主題組織發現,並附上內文來源引用及參考文獻章節。根據範圍與使用者目標,從四種輸出格式(快速簡報、研究摘要、比較分析、綜合報告)中選擇。使用內建範本建立與更新Notion頁面;直接連結來源,並在新資訊到來時追蹤變更。
official
rcsb-pdb-skill
openai
提交精簡的RCSB PDB請求以獲取核心元數據、Search API查詢及FASTA下載。當用戶需要簡潔的RCSB摘要時使用;保存原始JSON或…
official
pdf
openai
PDF 讀取、建立與驗證,支援視覺化渲染與程式化生成。使用 Poppler(pdftoppm)將 PDF 頁面渲染為 PNG,以便在交付前檢查版面、間距與排版。透過 reportlab 程式化生成 PDF,確保格式可靠;使用 pdfplumber 或 pypdf 提取文字與元資料。執行品質標準:無文字裁切、元素重疊、表格破損或渲染瑕疵;僅使用 ASCII 連字號,引用格式需具可讀性。使用...
official
test-coverage-improver
openai
Improve test coverage in the OpenAI Agents JS monorepo: run `pnpm test:coverage`, inspect coverage artifacts, identify low-coverage files and branches, propose…
official
playwright
openai
終端驅動的瀏覽器自動化,具備元素快照與互動式UI工作流程。透過playwright-cli包裝腳本運作(需npx),支援無頭模式與有頭模式以進行視覺化除錯。核心流程:開啟頁面、擷取快照以取得穩定元素參照、使用參照進行互動、在導航或DOM變更後重新擷取快照。包含表單填寫、點擊、輸入、多分頁管理、螢幕截圖/PDF擷取,以及流程除錯的軌跡記錄。元素參照(例如e3、e15)...
official
ukb-topmed-phewas-skill
openai
接受rsID、GRCh37或GRCh38輸入,解析為所需的GRCh38查詢,以獲取單一變異的簡潔UKB-TOPMed PheWAS摘要。當需要…時使用。
official
code-review-context
openai
模型可見上下文
official