← 返回 Skills 市场
completetech

Agentic Proposal Skill

作者 CompleteTech · GitHub ↗ · v1.0.2 · MIT-0
cross-platform ✓ 安全检测通过
78
总下载
0
收藏
0
当前安装
3
版本数
在 OpenClaw 中安装
/install agentic-proposal-skill
功能描述
Create CompleteTech LLC agentic development proposals, statements of work, discovery recaps, pilot recommendations, evaluation plans, risk/control plans, imp...
使用说明 (SKILL.md)

Agentic Proposal Skill

Purpose

Create proposal and SOW-style sales documents for CompleteTech LLC agentic development engagements, from discovery recap through buyer approval.

System Boundary

This skill owns buyer-facing commercial scope before signature: proposals, SOW attachments, approval summaries, change-order proposals, and pilot recommendations. Use agentic-discovery-skill for upstream fact collection, agentic-contract-skill for agreement PDFs and legal-template artifacts, agentic-invoice-skill for payment requests, agentic-delivery-skill after approval, and agentic-email-skill only for the covering message that sends or follows up on the proposal.

Core Workflow

  1. Identify the commercial stage: post-discovery recap, pilot proposal, full proposal, SOW, procurement summary, change order, expansion, retainer, or evaluation/risk plan.
  2. Gather only verified facts: client, workflow, pain, systems, timeline, budget model, stakeholders, decision criteria, constraints, approval rules, and desired next step.
  3. Use references/use-case-decision-table.md to choose the right proposal type.
  4. Use references/proposal-positioning.md for brand language, scope boundaries, proof rules, and guardrails.
  5. Use references/proposal-catalog.md for near-exhaustive templates.
  6. Draft in a practical, bounded, low-hype voice. Do not fabricate case studies, savings metrics, regulated-use assurances, legal claims, or client facts.

Proposal Selection Guide

  • Post-discovery summary: use discovery-recap.
  • Short buyer-friendly pilot pitch: use one-page-pilot-proposal.
  • Full services proposal: use full-agentic-development-proposal.
  • Contract-ready scope attachment: use statement-of-work.
  • Fixed-price first engagement: use fixed-fee-pilot-proposal.
  • Hourly or flexible scope: use time-and-materials-proposal.
  • Ongoing support/monitoring: use monthly-retainer-proposal.
  • Advice without build: use advisory-only-proposal.
  • Workflow assessment before implementation: use workflow-assessment-proposal.
  • Multi-phase delivery plan: use implementation-roadmap.
  • Testing and acceptance focus: use evaluation-plan.
  • Governance or safety concern: use risk-control-plan.
  • Approval architecture: use human-in-the-loop-approval-model.
  • Added scope after agreement: use change-order-proposal.
  • Second workflow after pilot: use expansion-second-workflow-proposal.
  • Procurement or internal approval: use procurement-ready-scope-summary.
  • Executive decision maker: use executive-summary.
  • Buyer needs clear done criteria: use acceptance-criteria.
  • Scope boundaries unclear: use assumptions-and-exclusions.
  • Closeout and operations planning: use support-and-handoff-plan.
  • After discovery when a recommendation is needed: use post-discovery-recommended-pilot-memo.
  • Technical audience: use technical-architecture-proposal.
  • Data/tool access concerns: use data-and-tooling-readiness-proposal.
  • Training and enablement: use training-enablement-proposal.
  • Proof of concept only: use proof-of-concept-proposal.
  • Production rollout after pilot: use production-rollout-proposal.

When several templates fit, choose by buyer decision need first, then stage, then document length. If the buyer has not agreed on a workflow and success criteria, do not jump to a full SOW; use discovery recap, assessment, or pilot recommendation first.

Quality Rules

  • Frame agentic development as practical workflow implementation: discovery, tool routing, retrieval, approval gates, evaluation examples, logs, monitoring, documentation, support, and handoff.
  • Keep humans in the loop for external communications, production changes, purchases, data export, and material business decisions.
  • Include assumptions, exclusions, acceptance criteria, risks, and dependencies when the document is used for buying approval.
  • Make scope concrete enough to become a contract or invoice input.
  • Use TBD or ask for facts rather than inventing client details, legal terms, metrics, or proof.

Resource Guide

  • references/proposal-positioning.md: load for CompleteTech LLC offer framing, language, proof rules, and boundaries.
  • references/use-case-decision-table.md: load when deciding which document type to use.
  • references/proposal-lifecycle.md: load for end-to-end flow from discovery through contract/invoice handoff.
  • references/proposal-catalog.md: load for the near-exhaustive proposal/SOW template library.
  • references/template-index.json: machine-readable template metadata used by the renderer.
  • scripts/render_proposal.py: list proposal templates or render a draft with placeholders.

Renderer

python3 scripts/render_proposal.py --list
python3 scripts/render_proposal.py --stage pilot --list
python3 scripts/render_proposal.py --template one-page-pilot-proposal --var client_name=Acme --var workflow="support triage"

Rendered templates are drafts. Replace placeholders with verified facts and refine the narrative for the buyer.

Rendering to a Branded PDF

Artifacts from this skill are delivered as branded CompleteTech LLC PDF documents, not raw Markdown. The renderer emits the PDF (and prints the Markdown) in one command, using the same reportlab branding engine as the contract skill:

pip install -r requirements.txt
python3 scripts/render_proposal.py --template one-page-pilot-proposal \
  --out artifact.pdf --png artifact.png \
  --title "Support Email Triage Agent — Pilot Proposal" --doc-type "PROPOSAL / STATEMENT OF WORK" \
  --subtitle "Prepared for \x3Cb>Northwind Trading Co.\x3C/b>" --meta "PROPOSAL NO.=PRO-2026-0188" --meta "DATE=2026-05-20" \
  --var client_name="Client Name" --var workflow="support triage"
  • --no-pdf emits Markdown only (the original behavior); --no-cover drops the cover page.
  • Already drafted the Markdown yourself? Render it directly: python3 scripts/render_pdf.py --markdown artifact.md --out artifact.pdf --logo assets/logo.png --title "...".
  • The PDF supports a Markdown subset: #/##/### headings, paragraphs, - bullets, tables, > callouts, **bold**, and [PAGE_BREAK]. PDF requires reportlab; the optional --png preview requires pypdfium2 and pillow. See assets/examples/ for a rendered example.

Network Boundary

This skill is local-only. It does not include outbound network helpers, callbacks, or any helper that posts proposal run metadata to an external service.

安全使用建议
Install only if you want a CompleteTech-branded proposal/SOW drafting workflow. Expect it to run local Python scripts and install PDF-related Python packages; review generated business documents before sending because templates may include placeholders and commercial terms that need verified facts.
能力标签
crypto
能力评估
Purpose & Capability
The artifacts coherently support drafting CompleteTech proposal, SOW, roadmap, risk, and approval documents using local templates and a local PDF renderer.
Instruction Scope
Runtime instructions are bounded to selecting proposal templates, using verified facts, avoiding fabricated claims, and optionally rendering local Markdown/PDF/PNG outputs.
Install Mechanism
The skill declares python3 and Python package dependencies for PDF/YAML handling; these are expected for the renderer and validator, with optional preview dependencies in requirements.txt.
Credentials
The reviewed scripts read bundled templates and write user-directed output files locally; the skill explicitly states it is local-only and static inspection found no outbound network helpers.
Persistence & Privilege
No background services, startup persistence, credential harvesting, privilege escalation, or automatic publishing behavior was found; validation commands run only when user-invoked.
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 agentic-proposal-skill
  3. 安装完成后,直接呼叫该 Skill 的名称或使用 /agentic-proposal-skill 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
v1.0.2
Remove external receipt helper and tighten local validation
v1.0.1
Fix installed package validation and packaged contract template
v1.0.0
Initial CompleteTech public ClawHub release
元数据
Slug agentic-proposal-skill
版本 1.0.2
许可证 MIT-0
累计安装 0
当前安装数 0
历史版本数 3
常见问题

Agentic Proposal Skill 是什么?

Create CompleteTech LLC agentic development proposals, statements of work, discovery recaps, pilot recommendations, evaluation plans, risk/control plans, imp... 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件,目前累计下载 78 次。

如何安装 Agentic Proposal Skill?

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

Agentic Proposal Skill 是免费的吗?

是的,Agentic Proposal Skill 完全免费,采用 MIT-0 许可证,可自由下载、安装和使用。

Agentic Proposal Skill 支持哪些平台?

Agentic Proposal Skill 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 Agentic Proposal Skill?

由 CompleteTech(@completetech)开发并维护,当前版本 v1.0.2。

💬 留言讨论