Skip to content
Get packs

Guide · Agent skill

How to use the Sales Engineer agent skill: RFP, POC and demo prep

Updated · written with AI from the project's own documentation and checked against our catalogue data

The Sales Engineer skill is a packaged set of instructions, templates and Python scripts that an AI agent can load to help with pre-sales engineering work: answering RFPs, comparing your product with competitors, and planning proof-of-concept (POC) evaluations. It is aimed at pre-sales or solutions engineers, sales teams and product marketers. After this guide you will know how to install it, how to ask for it in plain language, and what its scripts need from you.

Sales Engineer Toolkitalirezarezvani/claude-skills · MIT

Quick start

  1. 1Download the ZIP. Download the sales-engineer skill ZIP from this page.
  2. 2Unzip into a skills folder. Unzip it into ~/.claude/skills/ for personal use, or into .claude/skills/ inside a project.
  3. 3Or upload in the Claude apps. In the Claude apps, open Settings, go to Capabilities, then Skills, and upload the skill.
  4. 4Prepare your input JSON. Create JSON files describing your requirements, competitors and POC data, using assets/sample_rfp_data.json as the schema reference.
  5. 5Ask in plain language. Describe the task, such as an RFP response, competitor comparison or POC plan, and point the agent to your data file.
  6. 6Review the output. Check the coverage score, matrix or go/no-go recommendation and confirm it with a human before acting.

What does the sales-engineer skill make the agent do?

Once installed, the agent follows a five-phase pre-sales engineering process: discovery, solution design, demo prep, proof-of-concept (POC) planning, and proposal writing. For each phase the skill lists a checklist, the expected output and a validation checkpoint.

It can score how well your product covers an RFP's requirements, build a feature comparison against competitors, and generate a POC plan with a timeline, scorecard and a go/no-go recommendation. It also provides templates for demo scripts, POC scorecards and technical proposals.

  • RFP/RFI coverage analysis with gap identification and a bid/no-bid recommendation
  • Competitive feature comparison matrix with weighted scores, differentiators and vulnerabilities
  • POC plan with a default five-week breakdown, risk register and go/no-go framework
  • Templates: demo script, POC evaluation scorecard, technical proposal

How do you install the sales-engineer skill?

Download the ZIP file from this page. Then choose one of two routes. For a personal setup, unzip it into ~/.claude/skills/. For a single project, unzip it into the .claude/skills/ folder inside that project.

If you use the Claude apps instead, open Settings, go to Capabilities, and upload the skill under Skills.

  • Personal: unzip into ~/.claude/skills/
  • Project only: unzip into .claude/skills/ in your project folder
  • Claude apps: Settings → Capabilities → Skills, then upload the skill

What do you need before you start?

The scripts are written in Python 3 and use only the standard library, with no third-party packages visible in the source. You also need your own input JSON files describing your requirements, competitors and POC data. The bundled sample_rfp_data.json is an illustrative example, not a real customer dataset.

The scripts make no network calls, so they work on local files only.

  • Python 3 installed on the machine where the agent runs the scripts
  • Your own JSON files: RFP requirements, competitor feature data, POC data
  • Your own review: the output is a recommendation, not a final decision

How do you trigger the skill with a natural request?

You do not need to type a command. Describe the task in plain words and the agent will match it to the skill. The description covers requests about RFP responses, bids, proposals, competitor comparisons, feature matrices, POC planning, sales demo prep and pre-sales engineering.

Give the agent the relevant data file so it can run the scripts against it. Name the task type clearly so it picks the right phase.

  • "Help me write an RFP response for this bid. The requirements are in rfp_requirements.json."
  • "Build a feature matrix comparing our product with Competitor A and B using competitive_data.json."
  • "Plan a 5-week POC for this customer and tell me whether we should go or no-go."
  • "Prepare sales demo prep notes for the CTO, using the demo script template."

How do you run the scripts yourself?

The agent normally runs these for you, but you can call them directly. Each script has a human-readable mode and a JSON mode, and the RFP analyzer has a help flag.

The RFP analyzer scores coverage with weights of 3× for must-have, 2× for should-have and 1× for nice-to-have requirements. Coverage categories are Full (100%), Partial (50%), Planned (25%) and Gap (0%). A bid is recommended when coverage is above 70% with no more than three must-have gaps; a conditional bid applies at 50–70% coverage or two to three must-have gaps; otherwise it is no-bid.

  • Coverage analysis: python scripts/rfp_response_analyzer.py assets/sample_rfp_data.json
  • JSON output: add --format json
  • Help: python scripts/rfp_response_analyzer.py --help
  • Competitor matrix: python scripts/competitive_matrix_builder.py competitive_data.json --format json
  • POC plan: python scripts/poc_planner.py poc_data.json --format json
python scripts/rfp_response_analyzer.py assets/sample_rfp_data.json --format json

What checkpoints does the workflow enforce?

The skill sets validation gates between phases. Before moving to solution design, coverage must be above 50% and must-have gaps must be no more than three. Before a POC can convert, its scorecard must exceed 60% across functionality, performance, integration, usability and support. If the score is below 60%, the skill says to document the gaps and return to solution design.

Before a demo, the demo script must cover every must-have requirement flagged as a gap in the analysis output.

  • Phase 1 gate: coverage score above 50% and must-have gaps of 3 or fewer
  • Phase 2 gate: at least one strong differentiator per customer priority; otherwise escalate to the Product Team
  • Phase 4 gate: POC scorecard above 60% across all dimensions

What safety and data notes should you know?

The skill runs local Python scripts, so review them before use as you would any script you run on your machine. The scripts produce recommendations from simple weighted scores, not validated judgement. Final legal or pricing decisions must be reviewed by a human.

Keep customer data in your own files. The source states the skill does not use the network.

  • Read the scripts before running them on sensitive data
  • Treat bid/no-bid and go/no-go outputs as inputs to a human decision
  • Do not rely on the sample data for real scoring; replace it with your own inputs

What tips improve the results?

The scores are only as good as the input. Describe each requirement clearly, mark its priority honestly, and record competitor features with evidence rather than guesses. Use the templates in the assets folder to keep demo, POC and proposal documents consistent.

The references folder contains the RFP response guide, competitive positioning framework and POC best practices, which the agent can draw on for context.

  • Classify must-have, should-have and nice-to-have requirements carefully, since they change the weighting
  • Keep competitor data current and sourced
  • Use the POC scorecard template to record evidence for each rating
  • Ask the agent to explain any score before you act on it

What people use it for

  • Scoring how many of an RFP's must-have requirements your product covers before deciding to bid
  • Building a weighted feature comparison against two or three named competitors for a sales deck
  • Planning a five-week customer POC with setup, testing, evaluation and a go/no-go step
  • Preparing a demo script with talking points per stakeholder role using the demo template
  • Structuring a technical proposal with an implementation plan and risk mitigation section

Frequently asked questions

How do I install the sales-engineer agent skill?

Download the ZIP from this page and unzip it into ~/.claude/skills/ for personal use, or into .claude/skills/ in a project folder. In the Claude apps, you can instead upload it under Settings, then Capabilities, then Skills.

What do I need to run the sales-engineer skill?

You need Python 3, since the scripts use the standard library only. You also need your own input JSON files describing requirements, competitors and POC data. The scripts make no network calls.

Is the sales-engineer skill free?

The source gives the licence as MIT. Check the repository licence file for the full terms before you use or redistribute it.

What is the difference between the RFP analyzer and the competitive matrix builder?

The RFP analyzer scores how well your product covers an RFP's requirements and gives a bid/no-bid recommendation. The competitive matrix builder compares feature scores against competitors and identifies differentiators and vulnerabilities.

Why does the skill say my POC needs to score above 60%?

The skill sets 60% across functionality, performance, integration, usability and support as the threshold for POC conversion. Below that, it recommends documenting gaps and returning to solution design.

Can the agent make final pricing or legal decisions with this skill?

No. The source states the scripts produce recommendations from simple weighted scores, not validated judgement, so a human must review the results before any legal or pricing decision.

New agent skills, repos and AI tools — every week

The best new skills and open-source projects, the fastest-growing repos, and one free prompt. Five minutes to read. Unsubscribe anytime.

Our verdict

This skill suits pre-sales or solutions engineers, sales teams preparing competitive comparisons or customer POCs, and product marketers who want a structured bid/no-bid process. It is not useful if you have no real product or RFP to answer, because the scoring depends entirely on the data you provide. If you need validated pricing or legal judgement, use a qualified human reviewer or a dedicated tool instead.

Download the skill