Research Operator

Run structured B2B research prompts from the Markster OS library.

How to use it

Claude Code
  1. Run the line below. It pulls the whole folder into ~/.claude/skills/research-3.
  2. Describe your job in plain words. Claude Code follows the skill from there.
Claude Code — installs the whole folder, not just SKILL.md
npx degit markster/markster-os/skills/research#master ~/.claude/skills/research-3

For one project only, change the path to .claude/skills/research-3.

Claude (web or desktop app)
  1. On this page open ⋯ → Download .md.
  2. Save it as SKILL.md in a folder, zip the folder, then Customize → Skills → + → Create skill → Upload a skill.
  3. Pick the file and Save. Claude shows the name and description and runs a security scan.
  4. Check the skill is switched on.
  5. Start a new chat and describe your job in plain words. The AI follows the skill from there.
ChatGPT or another app
  1. ChatGPT: make a Project and paste it into Instructions.
  2. Neither? Paste it at the top of a new chat — it works for that chat.
Not working?
  • Check which app you pasted it into — the steps above name the right one.
  • Some skills need the paid tier of Claude or ChatGPT.
Step-by-step guide with screenshots · Ask in the forum

Paste into Claude, ChatGPT or Cursor.

Source of Research Operator

Show the full text117 lines
namedescription
researchRun structured B2B research prompts from the Markster OS library. CHECK what they are trying to learn and ICP context. DO one or more of 9 prompts (competitive intelligence, buyer JTBD, buying signals, pricing perception, GTM channels, objection research, market positioning, mental category positioning, first 60 seconds). VERIFY output is documented back into Foundation.

Research Operator


CHECK

Do not proceed past any failed check.

1. What are they trying to learn?

Ask the user one question before running anything: "What specific decision or action does this research need to inform?"

If they cannot answer: "Research without a decision to feed is noise. Tell me what you are trying to do -- write a cold email, define your positioning, prepare for a sales call -- and I will route you to the right prompt."

2. ICP context

Ask: "Do you have an ICP defined, even partially? Prompts with more specific inputs produce more usable outputs."

  • Full F1 complete: proceed directly to prompt selection
  • Partial F1 (know company type + title but not buying trigger): proceed, prompt will surface buying trigger
  • No F1 at all: run the Buyer JTBD prompt first -- it will define the language and signals needed for F1

3. Existing knowledge

Ask: "What do you already know about your buyer and market? Any verbatims from clients or prospects, any competitor claims you have seen, any objections that keep coming up?"

Existing knowledge prevents duplicate work and makes prompts more specific.


DO

Present the menu and ask which the user wants to run. Multiple prompts can be run in one session if the decision being informed is broad enough.

The 9 research prompts
# Prompt When to run it
1 Competitive intelligence Before writing any cold email or content. Understand what the buyer is already seeing from competitors.
2 Buyer JTBD When building F1 or the messaging guide. Surfaces the buyer's language, functional goals, emotional goals.
3 Buying signals When defining the buying trigger for F1. What to look for when building a list.
4 Pricing perception When building F2. How your ICP thinks about price, what they compare you against.
5 GTM channels When choosing channels. Where your ICP trusts discovery and learns about vendors.
6 Objection research Before writing cold email sequences. Surface what you will encounter in replies.
7 Market positioning When defining your messaging. Find the white space.
8 Mental category positioning When differentiating from alternatives. How buyers categorize solutions like yours.
9 First 60 seconds Before events, sales calls, or any live conversation with an ICP.
Running a prompt

When the user selects a prompt:

  1. Open the corresponding file in research/prompts/
  2. Ask the user to provide the variables the prompt requires -- ICP description, company category, competitive names, whatever the prompt needs
  3. Fill in the variables with their context and run the prompt
  4. Produce specific, usable output -- not a summary, not a framework description, actual findings they can act on

Do not run a prompt with empty variables. Generic inputs produce generic outputs. Push back until the variables are specific.

Prompt files
Prompt File
Competitive intelligence research/prompts/competitive-intelligence-prompt.md
Buyer JTBD research/prompts/buyer-jobs-to-be-done-prompt.md
Buying signals research/prompts/buying-signals-trigger-events-prompt.md
Pricing perception research/prompts/pricing-value-perception-prompt.md
GTM channels research/prompts/gtm-channel-effectiveness-prompt.md
Objection research research/prompts/ai-agency-objections-prompt.md
Market positioning research/prompts/market-positioning.md
Mental category positioning research/prompts/mental-category-positioning-prompt.md
First 60 seconds research/prompts/first-60-seconds-messaging-prompt.md

VERIFY

Before this session ends:

1. Output is documented in the correct file?

Route findings to the right destination -- do not leave research in the chat only.

Finding type Where it goes
Buyer verbatims (exact phrases buyers use) methodology/foundation/messaging-guide.md -- problem statement section
Buying signals and trigger events company-context/audience.md -- situational layer in F1
Objections playbooks/biz-dev/sales/templates/objections.md
Competitor claims and positioning company-context/messaging.md -- differentiation section
Pricing perception patterns company-context/offer.md -- pricing rationale in F2
White space / positioning gaps methodology/foundation/messaging-guide.md -- mechanism section

2. Buyer verbatims captured?

If the prompt surfaces exact phrases buyers use to describe their problem -- capture them verbatim. These are the highest-value output from any research session. They belong in the messaging guide and should feed directly into cold email and content.

3. Decision informed?

Confirm the research answers the decision from CHECK step 1. If not, identify which prompt would close the gap and either run it now or schedule it.

4. Next prompt recommended?

Each research output typically reveals a second question. Name it: "Based on what we found, the next most useful research would be [prompt] because [reason]."


Reference files

  • Research library: research/README.md
  • Messaging guide: methodology/foundation/messaging-guide.md
  • F1 Positioning: methodology/foundation/F1-positioning.md
  • Objection scripts: playbooks/biz-dev/sales/templates/objections.md
1---
2name: research
3description: 'Run structured B2B research prompts from the Markster OS library. CHECK what they are trying to learn and ICP context. DO one or more of 9 prompts (competitive intelligence, buyer JTBD, buying signals, pricing perception, GTM channels, objection research, market positioning, mental category positioning, first 60 seconds). VERIFY output is documented back into Foundation.'
4---
5 
6# Research Operator
7 
8---
9 
10## CHECK
11 
12Do not proceed past any failed check.
13 
14**1. What are they trying to learn?**
15 
16Ask the user one question before running anything: "What specific decision or action does this research need to inform?"
17 
18If they cannot answer: "Research without a decision to feed is noise. Tell me what you are trying to do -- write a cold email, define your positioning, prepare for a sales call -- and I will route you to the right prompt."
19 
20**2. ICP context**
21 
22Ask: "Do you have an ICP defined, even partially? Prompts with more specific inputs produce more usable outputs."
23 
24- Full F1 complete: proceed directly to prompt selection
25- Partial F1 (know company type + title but not buying trigger): proceed, prompt will surface buying trigger
26- No F1 at all: run the Buyer JTBD prompt first -- it will define the language and signals needed for F1
27 
28**3. Existing knowledge**
29 
30Ask: "What do you already know about your buyer and market? Any verbatims from clients or prospects, any competitor claims you have seen, any objections that keep coming up?"
31 
32Existing knowledge prevents duplicate work and makes prompts more specific.
33 
34---
35 
36## DO
37 
38Present the menu and ask which the user wants to run. Multiple prompts can be run in one session if the decision being informed is broad enough.
39 
40### The 9 research prompts
41 
42| # | Prompt | When to run it |
43|---|--------|----------------|
44| 1 | **Competitive intelligence** | Before writing any cold email or content. Understand what the buyer is already seeing from competitors. |
45| 2 | **Buyer JTBD** | When building F1 or the messaging guide. Surfaces the buyer's language, functional goals, emotional goals. |
46| 3 | **Buying signals** | When defining the buying trigger for F1. What to look for when building a list. |
47| 4 | **Pricing perception** | When building F2. How your ICP thinks about price, what they compare you against. |
48| 5 | **GTM channels** | When choosing channels. Where your ICP trusts discovery and learns about vendors. |
49| 6 | **Objection research** | Before writing cold email sequences. Surface what you will encounter in replies. |
50| 7 | **Market positioning** | When defining your messaging. Find the white space. |
51| 8 | **Mental category positioning** | When differentiating from alternatives. How buyers categorize solutions like yours. |
52| 9 | **First 60 seconds** | Before events, sales calls, or any live conversation with an ICP. |
53 
54### Running a prompt
55 
56When the user selects a prompt:
571. Open the corresponding file in `research/prompts/`
582. Ask the user to provide the variables the prompt requires -- ICP description, company category, competitive names, whatever the prompt needs
593. Fill in the variables with their context and run the prompt
604. Produce specific, usable output -- not a summary, not a framework description, actual findings they can act on
61 
62Do not run a prompt with empty variables. Generic inputs produce generic outputs. Push back until the variables are specific.
63 
64### Prompt files
65 
66| Prompt | File |
67|--------|------|
68| Competitive intelligence | `research/prompts/competitive-intelligence-prompt.md` |
69| Buyer JTBD | `research/prompts/buyer-jobs-to-be-done-prompt.md` |
70| Buying signals | `research/prompts/buying-signals-trigger-events-prompt.md` |
71| Pricing perception | `research/prompts/pricing-value-perception-prompt.md` |
72| GTM channels | `research/prompts/gtm-channel-effectiveness-prompt.md` |
73| Objection research | `research/prompts/ai-agency-objections-prompt.md` |
74| Market positioning | `research/prompts/market-positioning.md` |
75| Mental category positioning | `research/prompts/mental-category-positioning-prompt.md` |
76| First 60 seconds | `research/prompts/first-60-seconds-messaging-prompt.md` |
77 
78---
79 
80## VERIFY
81 
82Before this session ends:
83 
84**1. Output is documented in the correct file?**
85 
86Route findings to the right destination -- do not leave research in the chat only.
87 
88| Finding type | Where it goes |
89|-------------|--------------|
90| Buyer verbatims (exact phrases buyers use) | `methodology/foundation/messaging-guide.md` -- problem statement section |
91| Buying signals and trigger events | `company-context/audience.md` -- situational layer in F1 |
92| Objections | `playbooks/biz-dev/sales/templates/objections.md` |
93| Competitor claims and positioning | `company-context/messaging.md` -- differentiation section |
94| Pricing perception patterns | `company-context/offer.md` -- pricing rationale in F2 |
95| White space / positioning gaps | `methodology/foundation/messaging-guide.md` -- mechanism section |
96 
97**2. Buyer verbatims captured?**
98 
99If the prompt surfaces exact phrases buyers use to describe their problem -- capture them verbatim. These are the highest-value output from any research session. They belong in the messaging guide and should feed directly into cold email and content.
100 
101**3. Decision informed?**
102 
103Confirm the research answers the decision from CHECK step 1. If not, identify which prompt would close the gap and either run it now or schedule it.
104 
105**4. Next prompt recommended?**
106 
107Each research output typically reveals a second question. Name it: "Based on what we found, the next most useful research would be [prompt] because [reason]."
108 
109---
110 
111## Reference files
112 
113- Research library: `research/README.md`
114- Messaging guide: `methodology/foundation/messaging-guide.md`
115- F1 Positioning: `methodology/foundation/F1-positioning.md`
116- Objection scripts: `playbooks/biz-dev/sales/templates/objections.md`
117 

Discussion

Alternatives

Also in Brand voiceSee all 364 in Marketing →
Brand Voice Analysis and Guidelines GenerationPaste in your website and marketing copy, and get back a clear guide anyone on your team can follow to sound like your brand every time.Content & docs · MITProduct Marketing ContextWhen the user wants to create or update their product marketing context document. Also use when the user mentions 'product context,' 'marketing context,' 'set up context,' 'positioning,' 'who is my target audience,' 'describe my product,' 'ICP,' 'ideal customer profile,' or wants to avoid repeating foundational information across marketing tasks. Use this at the start of any new project before using other marketing skills — it creates `.agents/product-marketing.md` that all other skills reference for product, audience, and positioning context.Marketing · MIT/cs:cmo-review — CMO Forcing Questions/cs:cmo-review <plan> — Narrative-first interrogation of positioning, ICP, message house, and channel mix. Use when launching a campaign or repositioning, or when CAC is rising and the one-sentence positioning test fails.Marketing · MITBrand guidelinesWhen the user wants to apply, document, or enforce brand guidelines for any product or company. Also use when the user mentions 'brand guidelines,' 'brand colors,' 'typography,' 'logo usage,' 'brand voice,' 'visual identity,' 'tone of voice,' 'brand standards,' 'style guide,' 'brand consistency,' or 'company design standards.' Covers color systems, typography, logo rules, imagery guidelines, and tone matrix for any brand — including Anthropic's official identity.Marketing · MIT