Status: informative companion to GIDP Draft 0.1. Non-normative. Purpose: demonstrate that the same protocol primitives apply across materially different domains (the horizontal-layer hypothesis, GIDP 0.1 Appendix F.2), and give implementers concrete test scenarios. Scope note: this catalogue is illustrative and makes no claim to be exhaustive. It shows one or two generic cases per domain and does not analyse markets or business models.
Each case follows the same template so that the invariant structure is visible:
local (evaluation_only or never, GIDP 0.1 Section 10).The four classes of Conditional Interest (GIDP 0.1 Section 8) are abbreviated: PCD passive conditional demand, CAD confidential active demand, PCS private conditional supply, ICI interdependent conditional interest.
Classes: A = PCD, B = CAD.
Principal A: a senior executive, not looking. Principal B: a board considering succession without a public search.
Hidden interest: A would consider a CEO role; B would consider replacing its CEO.
Private conditions: A: minimum equity, excluded companies, earliest start date. B: timing, board context, current CEO’s awareness.
Why public discovery fails: A cannot post “open to CEO roles” without signalling to their current employer; B cannot post “CEO search” without destabilising the company.
GIDP interaction: both project executive_role · b2b_software · europe; the session confirms role, sector, scale band, geography and that each side’s private conditions are satisfiable; identities are revealed only after both Principals approve.
Resulting Opportunity: an introduction between two parties who never advertised. Neither equity floor nor succession timing was transmitted.
Classes: A = PCS, B = CAD.
Principal A: a founder-led company, not for sale. Principal B: an acquirer exploring targets discreetly.
Hidden interest: A would sell above a private threshold to a strategic buyer with management retained; B is building a position in a segment.
Private conditions: A: minimum valuation, acceptable buyer classes, founder transition terms. B: budget ceiling, strategic rationale, excluded competitors.
Why public discovery fails: a “for sale” signal damages A with customers, staff and competitors; an “acquiring” signal raises prices and alerts B’s competitors.
GIDP interaction: symmetric projections (strategic_transaction · enterprise_software · europe) hide direction; claims test valuation class and structure overlap; neither reservation value is transmitted.
Resulting Opportunity: a structure (“majority sale with founder retention”) and a compatible valuation band, handed to advisers. See GIDP 0.1 Appendix C.1 for the full walk-through.
Classes: A = CAD, B = PCD.
Principal A: a corporation with an undisclosed acquisition thesis. Principal B: a company that “would listen”.
Hidden interest: A’s thesis and budget; B’s openness.
Private conditions: A: target ARR band, maximum valuation, integration constraints. B: minimum price, buyer type.
Why public discovery fails: disclosing the thesis reveals strategy to competitors and moves prices.
GIDP interaction: as case 2, initiated from the buyer side; disclosure requests escalate from structure class to ARR band only after conditionally_compatible on valuation class.
Resulting Opportunity: a shortlist of two or three counterparties for A, none of which learned A’s identity or budget before consent.
Classes: A = ICI, B = CAD.
Principal A: an angel investor. Principal B: a startup raising quietly.
Hidden interest: A invests only if a qualified institutional lead commits; B does not want the market to know it is raising.
Private conditions: A: ticket size, required referral from a trusted connection, sector list. B: valuation, existing commitments.
Why public discovery fails: B cannot announce a round; A cannot broadcast “I follow leads”.
GIDP interaction: projections co-investment · ai · france; compatibility on sector, stage and ticket class; A’s conditional_on: institutional_lead is an unmet dependency that either resolves in session (B discloses a lead exists, at class session) or triggers dependency-driven expansion (GIDP 0.1 Section 19.2).
Resulting Opportunity: a conditional commitment, contingent on the lead, handed to the parties.
Classes: A = CAD, B = PCS.
Principal A: an enterprise whose CRM-resident agent knows deals are being lost for lack of capability X. Principal B: a startup that has X and lacks distribution.
Hidden interest: A’s product gap; B’s willingness to white-label.
Private conditions: A: commercial thresholds, lost-deal figures, roadmap. B: minimum contract value, exclusivity limits.
Why public discovery fails: A will not publish a product weakness; B will not signal it needs a partner.
GIDP interaction: projections expose capability classes (requires: capability X / provides: capability X) without motives; session confirms structure overlap (licensing, white-label) and threshold satisfiability.
Resulting Opportunity: a licensing or white-label discussion that neither party would have initiated publicly. Illustrates that participants need not be personal agents: an agent embedded in enterprise software already holds the relevant private state.
Classes: A = PCD, B = PCS.
Principal A: an organisation under contract until 2028, not running an RFP. Principal B: a supplier with unadvertised spare capacity.
Hidden interest: A would switch if integration, cost and risk thresholds are met; B would serve a new client above a size threshold.
Private conditions: A: current pricing, contract terms, switching-cost ceiling. B: capacity, floor price.
Why public discovery fails: an RFP signals dissatisfaction to the incumbent; advertised spare capacity weakens B’s pricing.
GIDP interaction: projections supplier_relationship · category · region; claims test volume class, integration class and timing window; cost thresholds stay evaluation_only.
Resulting Opportunity: a qualified pre-RFP conversation.
Classes: bilateral A = PCS, B = PCD; multi-party adds ICI.
Principal A: an owner who is not listing. Principal B: a buyer with a Standing Interest (“garden, under 20 minutes from X, below a private budget”).
Hidden interest: A would sell above a private price; B is not actively searching.
Private conditions: A: floor price, timing. B: budget ceiling, must-haves.
Why public discovery fails: listing commits A publicly; B’s conditions are too idiosyncratic and private for portals.
GIDP interaction: projections residential_property · region · size band; price class tested in session; identities and address at consent. In the multi-party variant (GIDP 0.1 Appendix C.4) a buyer’s requires: financing, occupancy and a bank’s and tenant’s provides … conditional_on … form a potentially satisfiable coalition that no bilateral pair could close.
Resulting Opportunity: a private viewing, or a feasible coalition for a commercial asset.
Classes: A = PCS, B = CAD.
Principal A: an expert who does not consult publicly. Principal B: a company that needs discreet advice on a sensitive project.
Hidden interest: A would advise under narrow conditions (topic, conflict-of-interest limits, rate); B cannot reveal the project.
Private conditions: A: excluded clients, rate floor. B: project nature, timeline.
Why public discovery fails: expert-network listings expose A’s availability; a public request exposes B’s project.
GIDP interaction: projections advisory · domain · language; conflict-of-interest exclusions evaluated locally against a disclosed counterparty class before identity is revealed.
Resulting Opportunity: a scoped advisory engagement.
Across all eight cases, the protocol mechanics are identical and only the vocabulary changes:
| Invariant | Person cases | Organisation cases |
|---|---|---|
Standing Interest with evaluation_only conditions |
equity floor, budget ceiling, excluded employers | reservation value, thresholds, lost-deal data |
| Discovery Projection that hides direction and identity | executive_role · sector · region |
strategic_transaction · domain · region |
| Coarse candidate retrieval | provider, community, network | registry, CRM ecosystem, bank |
| Compatibility on classes, not values | scale band, ticket class | valuation class, volume class |
| Progressive disclosure with stated purpose | timing, start date | structure class, ARR band |
| Consent before identity | session/principal_approval gate |
session/principal_approval gate (per Standing Interest) |
| Handoff | introduction, meeting | advisers, negotiation protocol |
This is the property GIDP 0.1 Appendix F.2 asks implementers to test: if the same implementation, with swapped vocabularies, runs cases 1, 2, 5 and 7, the horizontal-layer hypothesis survives. If each case needs a different state machine or disclosure model, it does not.
A demonstration that “agent A needs a Python developer and agent B knows one” is capability matching and proves nothing about GIDP. A convincing demonstration has two independently controlled Agents, each holding several Standing Interests, at least one real private constraint on each side, no public listing from which the Opportunity could have been found, staged retrieval, a compatibility result without secret disclosure, an explicit consent step, minimal disclosure, and an Opportunity at the end. See GIDP 0.1 Section 23.3.