← 返回 Skills 市场
qomob

pitch比稿技能

作者 qomob · GitHub ↗ · v1.0.1 · MIT-0
cross-platform ✓ 安全检测通过
93
总下载
0
收藏
0
当前安装
2
版本数
在 OpenClaw 中安装
/install pitchskill
功能描述
必赢逻辑引擎(Pitch Skill)— 专为广告/营销Agency的比稿竞标场景设计的AI影子智囊团。把资深策略总监脑子里的「玄学感悟」拆解为可计算的赢标逻辑。当用户需要在竞争性提案中赢下客户(多个供应商竞标、客户发RFP选Agency、评审团打分选方案)时使用此技能。6个Agent协作:Intake → In...
安全使用建议
This skill appears coherent for preparing competitive pitches and contains detailed, multi-step instructions that rely on external research and AIGC prompt generation. Things to consider before installing/using it: - Privacy: The Information Agent instructs the model to perform web searches and build decision-maker personas from public sources (LinkedIn, interviews). If you ask it to profile named individuals, confirm you are comfortable with public‑profile scraping and your platform's privacy rules. Avoid providing sensitive internal documents or private data unless you trust the runtime environment. - Internet & API behavior: The skill expects '联网搜索' and recommends AIGC services (Midjourney/DALL‑E). The skill itself does not require API keys, so it likely will produce search instructions and prompt packs for you to run. If you expect the platform to call external APIs automatically, verify whether that will happen and whether credentials would be needed/securely handled. - Ethical considerations: The SKILL.md repeatedly emphasizes psychological tactics ('压迫感', '心理统治') and even '制造执行力溢出的假象' (creating highly finished demos to project execution). While effective in competitive pitches, these instructions can border on manipulative or misleading practices. Decide whether your organization is comfortable using such tactics and ensure compliance with any client or legal standards (e.g., disclosing simulated demos vs. real deliverables). - Evidence provenance: Decision Agent requires evidence chains for WinRate scoring. If you expect high-confidence outputs, be prepared to supply or validate primary data (briefs, past campaign metrics). The skill will downgrade scores when evidence is lacking—useful, but verify the platform's ability to perform reliable searches. - Minor oddity: Some reference links include an absolute local file URI (file:///Users/jonki/...), which is harmless in this package but suggests the author used a local dev environment; nothing indicates exfiltration, but you may flag it for maintainers or ignore. If these considerations are acceptable and your runtime environment allows necessary web access (or you are comfortable running the produced prompts manually), the skill is coherent and functionally appropriate for its stated purpose.
功能分析
Type: OpenClaw Skill Name: pitchskill Version: 1.0.1 The Pitch Skill bundle is a sophisticated strategic tool designed for marketing agencies to prepare for competitive bidding (pitches). It implements a multi-agent workflow (Intake, Information, Strategy, Decision, Expression, and Delivery) to analyze client RFPs, profile decision-makers using OSINT (Open Source Intelligence) techniques, and generate persuasive pitch materials. The code and instructions (SKILL.md and agent-specific files) are consistently aligned with professional business intelligence and marketing strategy goals. No evidence of data exfiltration, malicious execution, or harmful prompt injection was found; the 'Information Agent' and 'Decision Agent' focus on legitimate competitive analysis and stakeholder mapping within a professional context.
能力评估
Purpose & Capability
Name/description promise (a multi-agent pitch assistant) aligns with the provided SKILL.md and the agents/*.md files. The skill is instruction-only, lists no binaries, env vars, or install steps—everything present (Intake/Information/Strategy/Decision/Expression/Delivery) is reasonable for a pitch-preparation tool.
Instruction Scope
The instructions explicitly tell agents to perform web searches (public search, LinkedIn/脉脉 lookups) and to produce AIGC prompt packs and pressure-test Q&A. That is consistent with 'Information Agent' and 'AIGC Demo' goals, but it means the agent will be expected to query external sources and reason about individuals (decision‑maker profiling). If the platform/agent runtime does not have internet access, parts of the workflow will be degraded. Also the language encourages strong persuasion tactics ('心理统治', '高压迫感', '制造执行力溢出的假象'), which is domain-coherent but has ethical implications (see guidance).
Install Mechanism
Instruction-only skill with no install spec and no code files to execute. Lowest install risk—nothing is downloaded or installed by the skill itself.
Credentials
Skill declares no required environment variables, credentials, or config paths. The runtime instructions reference external web searches and third-party AIGC tools (Midjourney/DALL‑E) but do not request API keys—appropriate for an instruction-only skill that outputs prompts rather than programmatically calling those services.
Persistence & Privilege
Skill does not request always:true and has normal invocation settings. It does not attempt to modify other skills or system-wide settings.
如何使用
  1. 确保已安装 OpenClaw(本地或 Docker 部署)
  2. 在对话框中输入安装命令:/install pitchskill
  3. 安装完成后,直接呼叫该 Skill 的名称或使用 /pitchskill 触发
  4. 根据 Skill 的参数说明提供必要输入,即可获得结构化输出
版本历史
v1.0.1
Pitch Skill v1.0.1 — changelog - 全新品牌定位为“必赢逻辑引擎/AI影子智囊团”,强调将资深策略总监的实战逻辑和“赢标确定性”计算公式化。 - 信息解构和输出路径进一步强调“心理穿透”、“竞品逻辑真空区推演”、“甲方决策语言翻译”、“胜率证据链”等甲方采购真实需求。 - 增强Information和Strategy Agent任务,包括需求解构、决策者深度画像、第一性原理推导与逻辑链自检。 - Expression Agent加入AIGC Demo及情绪引擎优化,输出高完成感Demo和逐段情感分析。 - 触发场景和适用范围更为细化,支持“选代理商”“换代理商”等典型竞标语境,并强化模糊需求自动触发。 - 所有核心执行环节和Checkpoint确认语句/机制优化为
v1.0.0
pitchskill 1.0.0 — 企业级赢标系统首个版本发布 - 提供专为广告/营销代理公司设计的AI决策式比稿系统,目标是提升赢标概率。 - 引入6大Agent协作流程:Intake → Information → Strategy → Decision → Expression → Delivery,覆盖从Brief结构化到Pitch Deck输出全链路。 - 强制三大规则:所有输出“决策语言化”、必须有竞品推演、必须有胜率评估,确保系统聚焦主目标。 - 支持断点续跑、用户校正机制、快速模式和自定义编排,灵活应对不同Pitch场景需求。 - 全面适配中英文输出和多语言Pitch场景,自动选择输出语言和结构模板。 - 明确场景适用范围,仅限竞争性提案、比稿、招标、RFP场景。
元数据
Slug pitchskill
版本 1.0.1
许可证 MIT-0
累计安装 0
当前安装数 0
历史版本数 2
常见问题

pitch比稿技能 是什么?

必赢逻辑引擎(Pitch Skill)— 专为广告/营销Agency的比稿竞标场景设计的AI影子智囊团。把资深策略总监脑子里的「玄学感悟」拆解为可计算的赢标逻辑。当用户需要在竞争性提案中赢下客户(多个供应商竞标、客户发RFP选Agency、评审团打分选方案)时使用此技能。6个Agent协作:Intake → In... 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件,目前累计下载 93 次。

如何安装 pitch比稿技能?

在 OpenClaw 或 Claude Code 对话框中运行命令「/install pitchskill」即可一键安装,无需额外配置。

pitch比稿技能 是免费的吗?

是的,pitch比稿技能 完全免费,采用 MIT-0 许可证,可自由下载、安装和使用。

pitch比稿技能 支持哪些平台?

pitch比稿技能 跨平台运行,可在任意部署了 OpenClaw / Claude Code 的环境中使用(cross-platform)。

谁开发了 pitch比稿技能?

由 qomob(@qomob)开发并维护,当前版本 v1.0.1。

💬 留言讨论