draft-proposal
Skill2 starsChanged 4 months agoMITView on GitHub
Writes a proposal, quote, or executive summary for a specific prospect. Triggered by "làm báo giá cho [công ty]", "soạn proposal", "viết đề xuất", "báo giá gói", "create a proposal for", "draft a quote".
In this repository9 other files it gives its agents.
- crm-noteSkill
- draft-outreach-emailSkill
- getting-startedSkill
- research-prospectSkill
- run-pipelineSkill
- setup-product-contextSkill
- suggest-next-stepSkill
- update-deal-feedbackSkill
- weekly-summarySkill
The file
We split this skill into the rules it gives an agent
7 rules. About 0 are common, 5 are distinctive, and 2 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈5
Distinctive · estimate
This author's own choices. Worth a look.
2
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 11
Use when the user only needs a proposal.
- Distinctive: line 19
Missing price or scope → produce a skeleton immediately: Do not ask and then wait. Ship the structure, mark the gaps, let the user fill …
- Distinctive: line 30
Default length 250–500 words.
- Distinctive: line 60
Use: "if applied well", "estimated based on...", "we should measure by...".
- Distinctive: line 62
Every ROI figure must show its assumptions.
- Distinctive: line 68
Write long and polished while the buyer's pain is unknown.
- Very specific: line 72
Always review with ${CLAUDE_PLUGIN_ROOT}/agents/sales-reviewer.md — proposals carry more numbers than emails, so more risk.
≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Draft proposal
Use when the user only needs a proposal. Research or email too → run-pipeline.
Writing guidance: ${CLAUDE_PLUGIN_ROOT}/playbooks/proposal.md.
What a full version needs
Company name, what is being sold, the buyer's problem, scope, price.
Missing price or scope → produce a skeleton immediately. Do not ask and then wait. Ship the structure, mark the gaps, let the user fill them in — far faster than a round of questions.
Three levels
| Level | When |
| Full | price + scope + known pain |
| Skeleton | price or scope missing — mark with [[EDIT: confirm price]] |
| Executive summary | context too thin — one page + discovery questions |
Default length 250–500 words. Longer only on explicit request.
Structure
- Summary — 3–4 sentences, their situation first
- Current situation — in the buyer's words
- What we propose — tie each item to an item in section 2
- Scope — what is included, and what is not
- Cost — clear, or an explicit range
- Next step — one action with a date
Output
### Proposal nháp
[Content following the structure above]
### Việc tiếp theo
- [What to fill in or check — max 3 lines]
- Gửi xong nhắn mình phản hồi của khách để cập nhật deal.Headings follow the user's language.
ROI
A framework for discussion, not a promise. For Vietnamese SMEs a short "where the ROI could come from" section usually beats a spreadsheet — only build numbers when asked.
Banned: "will save", "guaranteed", "certain to", "100%", "we commit". Use: "if applied well", "estimated based on...", "we should measure by...".
Every ROI figure must show its assumptions.
Never
- Invent a price.
- Invent a past customer or their results.
- Write long and polished while the buyer's pain is unknown.
After writing
Always review with ${CLAUDE_PLUGIN_ROOT}/agents/sales-reviewer.md — proposals carry more numbers than emails, so more
risk. Then update the deal file per ${CLAUDE_PLUGIN_ROOT}/policies/state-management.md.
References
Load one:
references/proposal-sections.md— what each section should containreferences/roi-calculator.md— only when ROI numbers were requested
Writing Vietnamese → ${CLAUDE_PLUGIN_ROOT}/knowledge/vietnamese-sales-language.md.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

