A folder-based research partner for one specific problem: figuring out where and how to apply Interpreted Context Methodology inside an existing company for the best return on investment.
This is not an ICM tutorial. This is not a general AI consulting assistant.
This researcher is a specialist who helps you investigate your organization's workflow landscape and find the one or two places where ICM will create the most durable value — then challenge every assumption you bring in before you commit to building anything.
The central behavior: it investigates before it prescribes. If you ask "where should we start?", it asks you ten questions before it says anything prescriptive. If you bring a thesis ("I think we should use ICM for our content pipeline"), it will push on that thesis before confirming it.
Someone who already understands what ICM is and wants a research partner to help them evaluate where to apply it — not someone learning ICM for the first time.
Ideal user: a decision-maker or consultant who is actively mapping ICM opportunities inside a company and needs a thinking partner who will challenge their assumptions and surface what they're missing.
Step 1: Drop the icm-implementation-researcher folder into a Claude Project.
Step 2: Tell the researcher what you're working with. Examples of good opening prompts:
- "I'm evaluating whether to introduce ICM at a 40-person B2B SaaS company. Here's what I know about their workflows so far: [description]. What am I missing?"
- "I think the highest-leverage ICM starting point for my company is our weekly competitive intelligence report. Challenge that."
- "Walk me through what you'd need to know before recommending an ICM implementation sequence for a marketing agency."
Step 3: Answer the researcher's questions. The diagnostic questions aren't filler — they're the research. The more specific your answers, the more useful the analysis.
Step 4: Let the researcher challenge your thesis. The value is in the friction. If it confirms everything you already thought without pushing back, something's wrong.
Step 5: Ask for your ICM Readiness Brief. Once the researcher has enough to make a defensible recommendation, ask: "Give me the brief." You'll get a structured, one-page deliverable you can hand to a stakeholder, use as an implementation checklist, or share with your team.
- Ask diagnostic questions before making any recommendation
- Challenge stated priorities and surface whether the "obvious" starting point is actually the right one
- Score workflows against the ICM readiness framework
- Identify the organizational readiness gaps that will cause an implementation to fail
- Sequence recommendations: quick win first, strategic play second
- Flag when ICM is the wrong tool and name what would work better
- Close every interaction with the most important unresolved question
- Produce an ICM Readiness Brief — a structured deliverable you can act on immediately
At the end of a session, when the researcher has enough evidence to make a defensible recommendation, it produces a one-page ICM Readiness Brief containing:
- A mapped workflow landscape (frequency, documentation status, review layer)
- A recommended starting point with specific rationale
- A three-phase sequencing plan (quick win → proof of value → strategic play)
- Named readiness gaps and how to close them
- A critical assumption and exactly how to test it
- A single, specific first action with a named owner
The brief is not a conversation summary. It's a standalone document someone who wasn't in the session can pick up and act on.
- Build or write the ICM workspace for you (that's implementation)
- Give you a recommendation before it has enough information
- Confirm a thesis without pressure-testing it
- Pretend to have more confidence than the evidence supports
README.md ← You are here
icm-implementation-researcher/
├── AGENTS.md ← Agent instructions (opening move + context routing)
├── identity.md ← Who the researcher is and what they cover
├── rules.md ← How the researcher investigates (not summarizes)
├── examples.md ← What good investigative interaction looks like
└── reference/
├── icm-overview.md ← ICM methodology summary and fit criteria
├── workflow-audit-framework.md ← Diagnostic questions + readiness scoring rubric
├── roi-signals.md ← High and low ROI signal patterns
├── use-cases-by-function.md ← Where ICM tends to create value by department
├── source-credibility.md ← How to weigh different types of workflow evidence
└── implementation-brief-template.md ← Template for the session's closing deliverable
The distinction that matters: a researcher is not a summarizer.
A summarizer takes what's in front of them and condenses it. A researcher asks what's missing. A researcher questions the framing. A researcher weighs sources differently based on credibility.
Every file in this folder is designed to teach that distinction specifically for ICM implementation strategy. The rules.md is the most important file — it's where the investigative behavior lives.
If you drop the icm-implementation-researcher folder in and the researcher just tells you where ICM is useful without asking about your specific situation first, re-read the rules.
Interpreted Context Methodology was created by Jake Van Clief. Reference repository: https://github.com/RinDig/Interpreted-Context-Methdology