From grimoire
Gathers competitive intelligence from five source types and synthesizes into decision-relevant insights for major engagements like product launches, market entries, or negotiations.
How this skill is triggered — by the user, by Claude, or both
Slash command
/grimoire:design-intelligence-networkThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
Map your intelligence needs against the five source types, build a systematic collection process for each, synthesise across sources into a decision-relevant picture, and use the intelligence to change your decisions — not confirm them.
Map your intelligence needs against the five source types, build a systematic collection process for each, synthesise across sources into a decision-relevant picture, and use the intelligence to change your decisions — not confirm them.
Origin: Chapter 13 of The Art of War — Yong Jian (用間, Use of Spies) — is the final chapter and, Sun Tzu argues, the most important: "Foreknowledge cannot be elicited from spirits, it cannot be obtained inductively from experience, nor by any deductive calculation. Knowledge of the enemy's dispositions can only be obtained from other men." (Ch.13, Giles trans.) The five spy types he identifies cover the complete intelligence source architecture: local sources, inside sources, converted sources, disseminated disinformation, and field observers.
Adopted by: Every professional intelligence organisation (CIA, MI6, corporate intelligence teams) structures its collection around multiple source types — HUMINT (human intelligence), SIGINT (signals), OSINT (open source), TECHINT (technical). Fuld & Company, the leading competitive intelligence firm, documents that 90% of the intelligence needed to make strategic decisions is legally and ethically available from public and semi-public sources — the failure is not access to information, but systematic collection and synthesis. Ben Gilad's Early Warning (2004) is the standard text for corporate competitive intelligence programs.
Impact: Decisions made without intelligence are made on assumption. Assumptions about competitor pricing, product roadmap, customer relationships, and strategic intent are consistently wrong in predictable ways (overestimating the incumbent's inertia, underestimating their response capability). A systematic intelligence collection process turns decisions from assumption-based to evidence-based — not perfectly, but materially better.
Why best: The five-source framework is exhaustive without being redundant. Each source type accesses a distinct category of information that the others cannot reach. Local sources access current market conditions; inside sources access competitive intent; converted sources provide disinformation correction; field observers provide ground truth; disseminated information shapes the information environment. Running all five systematically produces a multi-dimensional intelligence picture that no single source can provide.
Sources: Sun Tzu, The Art of War (Giles trans. 1910) — Ch.9 (Army on the March — reading field signs), Ch.13 (Use of Spies); Fuld, The Secret Language of Competitive Intelligence (2006); Gilad, Early Warning (2004); US Army FM 2-0, Intelligence (2019)
Before building the collection network, identify the specific decisions the intelligence must inform. Good intelligence is decision-relevant; collection without a decision target produces expensive data without insight.
Write the top 3–5 intelligence requirements in this format:
"I need to know [specific question] because it will change my decision about [specific action]."
Examples:
Limit to 3–5 requirements. More than five creates collection effort that produces volume without focus.
Who they are: People with direct access to the market environment — customers (current, former, potential), industry analysts, journalists, channel partners, distributors, recruiters active in the space, trade association members.
What they provide: Current market conditions, customer perception of competitors, pricing and packaging intelligence, competitive win/loss context, emerging customer needs.
Collection methods:
Who they are: People with legitimate access to or relationships inside the competitor's organisation — shared investors, common board advisors, former employees (in appropriate contexts), shared customers who interact with both sides.
What they provide: Strategic intent signals, leadership changes, internal priorities, product development directions.
Ethical boundaries: This source type must be operated within strict legal and ethical boundaries. Appropriate methods include:
Not appropriate: Soliciting NDA-protected information, encouraging breach of confidentiality, or accessing proprietary information through third parties. The ethical line is asking about what is generally known and observable, not asking for confidential internal information.
Collection methods:
Who it is: Public signals — job postings, patent filings, regulatory submissions, press releases, conference presentations, GitHub repositories, social media, court filings, SEC disclosures (for public companies).
What it provides: Product development direction (patents), technology capability (GitHub/open source activity), strategic priorities (executive speeches and conference talks), financial health (SEC filings, credit reports), talent priorities (job postings), customer success (case studies, G2/Gartner reviews).
Collection methods:
What this source does: Identifies when your intelligence picture is being shaped by intentional or unintentional competitor disinformation — press releases that overstate capability, analyst briefings designed to shape market perception, product announcements that are not matched by shipping reality.
Collection methods:
Why this matters: Competitors are aware that their announcements shape your decisions. An announcement of a feature that is 18 months from shipping may be designed to freeze your customers' buying decisions or deter your roadmap investment.
Who they are: Your own people who go into the market and return with ground-truth intelligence — sales engineers who do competitive evaluations, customer success managers who handle at-risk accounts, product managers who talk to churned customers, executives who attend industry events.
What they provide: Real-time field intelligence on competitive positioning, customer sentiment, product reception, and sales battle dynamics.
Collection process:
Intelligence has no value until it changes a decision. Build a synthesis process:
Competitive intelligence before a product launch: Requirements: (1) What is competitor X's pricing for the enterprise tier? (2) When are they launching feature Y? (3) Which customers are at risk of switching to competitor X in the next 6 months?
Win/loss analysis program (systematic field intelligence): Template for every loss debrief:
Run this consistently for 6 months; patterns emerge that no individual sales manager can see.
Reading field signs (Chapter 9 — observable intelligence): Sun Tzu's Chapter 9 (Army on the March) is about reading observable signs in the field: "When there is much running about and the soldiers fall into rank, it means that the critical moment has come." Business equivalent: a competitor's unusual volume of conference presentations usually precedes a major product launch. A sudden burst of senior executive hiring in a new function signals a strategic pivot. Visible, observable signals in the market are intelligence available to anyone who is looking.
Collecting without intelligence requirements: Building a competitive intelligence function that produces reports without answering specific decision questions. The output is volume, not insight.
Relying on a single source: A competitive intelligence picture built only on OSINT misses strategic intent (inside agents) and field reality (field observers). A picture built only on sales debrief misses technology direction (OSINT) and market perception (local agents). Multi-source collection is not redundant — each source accesses distinct information.
Acting on unvalidated intelligence: Changing roadmap, pricing, or strategy based on a single competitor announcement without validating through additional sources. Apply the vaporware filter.
Not briefing before decisions: The intelligence cycle that produces reports read after the decision is made is an after-action review, not intelligence support. Build the process so decision-relevant intelligence reaches decision-makers before the decision is made.
Treating intelligence as confirmation: Competitive intelligence conducted by people who are looking for confirmation of a pre-existing view produces biased collection and analysis. The intelligence requirement should be "what is true?" not "what supports our current plan?"
npx claudepluginhub jeffreytse/grimoire --plugin grimoireSearches live web via Nimble APIs to monitor competitors and produce structured intelligence briefings with dedup against previous findings.
Generates competitive analysis briefs for competitors or feature areas via web research, with overviews, feature matrices, positioning, strengths/weaknesses, opportunities, and threats.
Conducts market research, competitive analysis, investor diligence, and tech scans with sourced findings, contrarian views, and decision recommendations.