{"schema_version":1,"research_id":"eoa_inverse_innovation_exp06_external_evaluation_20260803","source_assessment_id":"representation_independent_interface_contract__sociology_anthropology:P3:v0","cell_id":"representation_independent_interface_contract__sociology_anthropology","search_queries":["mutual aid groups digital tools spreadsheets privacy coordination referrals research","mutual aid coordination technology challenges duplicate requests status tracking study","Open Referral Human Services Data Specification referrals status workflow standard official","211 mutual aid referral interoperability standard API case management","official closed loop referral platform referral status accepted in progress completed Unite Us","findhelp referral status closed loop official platform","mutual aid platform request matching status coordination autonomous groups software","mutual aid networks technology infrastructure coordination study spreadsheets privacy","site:mutualaidhub.org technology tools coordination mutual aid groups","site:mutualaiddisasterrelief.org data privacy coordination tools mutual aid","mutual aid organizers survey digital tools coordination platform autonomy research PDF","mutual aid network federation coordination referrals autonomous groups software API","mutual aid organization needs better coordination technology volunteer requests spreadsheet named network","mutual aid group software request system open source dispatch GitHub","Mutual Aid Hub API groups directory official","mutual aid federation member groups referrals coordination","\"Mutual Aid Praxis Aligns Principles and Practice\" DOI","site:journals.sagepub.com \"Mutual Aid Praxis Aligns\"","\"Mutual Aid Praxis Aligns Principles and Practice in Grassroots COVID-19 Responses Across the US\""],"sources":[{"source_id":"S1","title":"“It Doesn’t Actually Feel Very Mutual:” How Technology Impacts the Values of Mutual Aid Groups in Practice","publisher":"Association for Computing Machinery, CHI 2025","url":"https://www.tonyanguyen.com/publications/mutual_aid.pdf","source_class":"PRIMARY_RESEARCH","publication_date":"2025-04-26","accessed_at":"2026-08-03","claims_supported":["Interviews with 12 US mutual-aid volunteers found that groups used heterogeneous digital tools for logistics and information management.","Existing technologies improved efficiency but created tensions involving security, accountability, access, governance, solidarity, and exclusion.","Uniform, scale-oriented systems can conflict with mutual aid's decentralized, relational, and hyperlocal values."]},{"source_id":"S2","title":"Mutual Aid Praxis Aligns Principles and Practice in Grassroots COVID-19 Responses Across the US","publisher":"Kennedy Institute of Ethics Journal; indexed by PubMed/NCBI","url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC10927022/","source_class":"PRIMARY_RESEARCH","publication_date":"2023-06-01","accessed_at":"2026-08-03","claims_supported":["Interviews with 15 organizers found mutual-aid networks coordinating groceries and other basic-needs support through online platforms, forms, messaging, email, and telephone channels.","Networks valued decentralized structure, equity in decision-making, solidarity, and mutuality but struggled to align those principles with practice.","Digital tools were useful but insufficient; coordination still depended substantially on human infrastructure."]},{"source_id":"S3","title":"Keeping Each Other Safe When Virtually Organizing Mutual Aid","publisher":"Electronic Frontier Foundation","url":"https://www.eff.org/deeplinks/2020/03/keeping-each-other-safe-when-virtually-organizing-mutual-aid","source_class":"AUTHORITATIVE_SECONDARY","publication_date":"2020-03-19","accessed_at":"2026-08-03","claims_supported":["Mutual-aid organizers used spreadsheets, forms, social posts, and private messages to coordinate support.","Organizers approached EFF for privacy and security guidance.","Overbroad collection and weak permissions can disclose sensitive information; EFF recommends data minimization, restricted access, encryption, and deliberate trust boundaries."]},{"source_id":"S4","title":"mutua — infrastructure for community care","publisher":"mutua","url":"https://www.mutua.nyc/","source_class":"OFFICIAL_ORGANIZATION_DATA","publication_date":"2025","accessed_at":"2026-08-03","claims_supported":["A named mutual-aid technology initiative is building infrastructure for mutual-aid networks and inviting organizers, developers, and community members to participate.","Its stated requirements include community control over information sharing, privacy, participatory design, and transparent sharing protocols.","Its FireAid case describes converting a community spreadsheet into an interface used during the 2025 Los Angeles fires."]},{"source_id":"S5","title":"Open Referral Data Specifications 3.0 — About and normative scope","publisher":"Open Referral Initiative","url":"https://docs.openreferral.org/en/latest/index.html","source_class":"STANDARD","publication_date":"n.d.; live documentation accessed 2026-08-03","accessed_at":"2026-08-03","claims_supported":["HSDS defines a normative data model, API reference, identifiers, profiles, serialization, and conformance rules for exchanging community-resource information.","The ecosystem includes reusable open-source applications and HSDS-compatible APIs.","HSDS principally standardizes resource-directory information, not the proposed provider-controlled commitment-transition semantics."]},{"source_id":"S6","title":"Unite Us Closed-Loop Referrals","publisher":"Unite Us","url":"https://uniteus.com/products/closed-loop-referral-system/","source_class":"COMMERCIAL_FIRST_PARTY","publication_date":"n.d.; accessed 2026-08-03","accessed_at":"2026-08-03","claims_supported":["An established product already supports secure electronic referrals among community partners, ongoing referral-status tracking, integrations, and resolved/unresolved outcomes.","The product demonstrates practical implementation of stateful, cross-organizational social-care referrals.","Its documented approach is a shared vendor platform and trusted network rather than a public conformance contract for independently operated mutual-aid adapters."]},{"source_id":"S7","title":"How It Works","publisher":"Coordinated Assistance Network","url":"https://portal.canportal.org/how-its-works","source_class":"COMMERCIAL_FIRST_PARTY","publication_date":"n.d.; site copyright 2021","accessed_at":"2026-08-03","claims_supported":["CAN offers cross-provider request coordination, customizable controls, permissions, case management, referral collaboration, and reduced duplicate requests.","Its workflow includes explicit Open, Review, and Claimed behavior with automatic rollback after a time limit.","CAN is close functional prior art but centralizes applicant history and partner access rather than hiding each member group's representation behind independently tested adapters."]},{"source_id":"S8","title":"Shareish (Share & Cherish): an open-source, map-based, web platform to foster mutual aid","publisher":"Association for Computing Machinery; University of Liège repository","url":"https://orbi.uliege.be/handle/2268/302644","source_class":"PRIMARY_RESEARCH","publication_date":"2023-05-29","accessed_at":"2026-08-03","claims_supported":["Researchers implemented an open-source mutual-aid platform supporting localized offers and requests, messaging, search, and notifications.","The system is intended to be replicable and modifiable by grassroots movements seeking autonomy.","It demonstrates technical feasibility of community-controlled mutual-aid software but not federation-wide behavioral conformance across heterogeneous local systems."]}],"problem_evidence":{"support":"MODERATE","rationale":"Multiple independent studies and EFF directly document heterogeneous tools, human-intensive coordination, privacy exposure, and value conflicts in mutual-aid technology. CAN additionally treats duplicate requests as a practical coordination problem. However, no source measures how often federation dispatchers specifically misclassify accepted, released, blocked, or completed commitments after local tool changes, so the candidate's precise failure mode remains unquantified.","source_ids":["S1","S2","S3","S7"]},"stakeholder_evidence":{"support":"MODERATE","rationale":"mutua is an identifiable, fiscally sponsored initiative explicitly seeking privacy-preserving, community-controlled sharing protocols for mutual-aid networks, while EFF reports direct requests from organizers for security guidance. These establish a credible adopter class and expressed adjacent need, but no federation or member group is documented as requesting or authorizing this exact commitment contract.","source_ids":["S3","S4"]},"prior_art":{"proximity":"SUBSTANTIAL_COLLISION","closest_analogues":[{"name":"Unite Us Closed-Loop Referral System","similarity":"Supports secure cross-organization referrals, detailed status and outcome tracking, partner workflows, and system integrations—the central operational surface proposed here.","remaining_difference":"The public materials describe a vendor-operated trusted network, not member-owned adapters independently certified against an open behavioral oracle with explicit provider-only accept/release authority.","source_ids":["S6"]},{"name":"Coordinated Assistance Network","similarity":"Coordinates assistance requests among multiple providers, reduces duplicates, uses permissions and explicit claim/open transitions, and automatically rolls claims back after a deadline.","remaining_difference":"It centralizes applicant history and workflow in one platform; the proposal instead claims representation independence across autonomous local tools and minimizes federation access to internal records.","source_ids":["S7"]},{"name":"Open Referral HSDS","similarity":"Provides a governed, machine-readable human-services data/API standard with profiles and conformance requirements, enabling heterogeneous systems to exchange community-resource information.","remaining_difference":"HSDS standardizes service-directory data and API conformance, not the full lifecycle, authority, idempotency, withdrawal, release, notification, and side-effect laws of a voluntary aid commitment.","source_ids":["S5"]},{"name":"Shareish","similarity":"Implements open-source, replicable mutual-aid request/offer infrastructure designed for grassroots autonomy.","remaining_difference":"It is an application that groups can deploy, not a federation protocol or shared black-box conformance suite for substitutable adapters over different local workflows.","source_ids":["S8"]}],"distinctive_claim_remaining":"For member-authorized mutual-aid groups that retain heterogeneous local tools, independently implemented adapters conforming to a minimal, authority-scoped Aid Commitment state machine will produce the same public outcome on held-out request histories—including retries, reordering, withdrawal, blockage, release, expiry, and invalid completion—without federation access to rosters or deliberation. The claim is falsified by any contract-conforming adapters that disagree on a held-out public outcome, any ordinary scenario requiring hidden-state inspection, protected data leakage, or member-review findings that the state model converts negotiation into assignment.","confidence":"HIGH"},"implementation_evidence":{"support":"MODERATE","rationale":"Open Referral demonstrates schemas, profiles, APIs, and conformance; Unite Us and CAN demonstrate deployed cross-organizational referral state tracking; Shareish demonstrates community-controlled mutual-aid software. These make a synthetic state machine, adapters, idempotent transition API, and black-box tests technically conventional. Feasibility remains uncertain for mapping relational or multi-provider practices, governance of semantic changes, notification failures, identity/authentication, accessibility, privacy-law applicability, operational support, and integration with actual group tools. A fictional-data tabletop avoids immediate legal exposure but cannot establish live safety or authority feasibility.","source_ids":["S1","S3","S5","S6","S7","S8"]},"scores":{"meaningful_impact":{"score":3,"rationale":"The intervention could reduce duplicate or stale routing and unnecessary disclosure, but the frequency and consequence distribution of the exact status-translation failures are not measured.","source_ids":["S1","S3","S7"]},"stakeholder_pull":{"score":3,"rationale":"A named initiative seeks community-controlled mutual-aid infrastructure and sharing protocols, but no adopter has committed to this specific contract or pilot.","source_ids":["S4"]},"incremental_advantage":{"score":3,"rationale":"Provider-owned adapters and a shared behavioral oracle could preserve more local autonomy than centralized referral platforms while specifying more lifecycle semantics than HSDS; comparative performance is untested.","source_ids":["S5","S6","S7"]},"distinctiveness_plausibility":{"score":2,"rationale":"Closed-loop referrals, status transitions, permissions, interoperability APIs, and conformance are already substantial prior art. The remaining distinction is their combination as an open, member-governed adapter contract for mutual aid.","source_ids":["S5","S6","S7","S8"]},"technical_implementability":{"score":4,"rationale":"State machines, stable transition identifiers, adapters, APIs, and black-box tests are standard engineering work, with multiple working analogues. Semantic mapping and identity boundaries are the harder portions.","source_ids":["S5","S6","S7","S8"]},"adoption_authority_feasibility":{"score":3,"rationale":"Member authorization and requester withdrawal fit documented mutual-aid commitments to decentralization and equity, but a federation-wide steward, decision process, and participating organizations have not been secured.","source_ids":["S1","S2","S4"]},"evidence_readiness":{"score":4,"rationale":"The proposed fictional tabletop is bounded, reversible, comparator-based, and technically buildable; it still requires recruitment and field participation from authorized representatives.","source_ids":["S4","S5","S8"]},"safety_net_benefit":{"score":4,"rationale":"Mutual-aid networks serve basic-needs gaps and marginalized communities, and privacy-minimizing coordination could directly benefit people exposed by current ad hoc tools. Benefits are conditional on preserving relational and non-hierarchical practice.","source_ids":["S1","S2","S3"]},"scalability":{"score":3,"rationale":"A common contract can reuse one oracle across adapters, but onboarding, semantic governance, accessibility, and support burdens may scale with every group and workflow variation.","source_ids":["S1","S4","S5"]}},"score_confidence":"MODERATE","costs":{"first_evidence":{"band_2026_usd":"10K_TO_50K","scope":"Two-week non-live tabletop with three group representatives, a requester advocate, a dispatcher, facilitation, participant compensation, 16 fictional histories, two minimal adapters, a router, generated transition tests, and a short leakage review.","confidence":"MODERATE","assumptions":["Approximately 4–8 person-weeks of engineering, facilitation, and analysis.","Existing spreadsheet/chat sandboxes and open-source test tooling are used.","No live personal data, production security certification, procurement, or legal opinion is included.","This is a resource-equivalent planning estimate; the sources do not publish prices for this experiment."],"source_ids":["S4","S5","S8"]},"initial_deployment_startup":{"band_2026_usd":"50K_TO_250K","scope":"Governed v1 contract, production-grade reference router, authentication and authorization design, audit and notification semantics, two to five hardened adapters, accessibility work, threat modeling, documentation, and independent security/privacy review.","confidence":"LOW","assumptions":["A small federation with at most five initial groups.","Existing group tools offer usable APIs or controlled export mechanisms.","No replacement of groups' internal systems.","Jurisdiction-specific counsel and extensive legacy integration could move the cost upward."],"source_ids":["S3","S5","S6"]},"operational_launch":{"band_2026_usd":"250K_TO_1M","scope":"Partner onboarding, adapter implementation and validation across roughly 5–15 groups, training, help desk, observability, incident response, governance meetings, migration/rollback exercises, and a staged non-emergency then live launch.","confidence":"LOW","assumptions":["At least two full-time technical/operations roles during launch plus paid integration and community-governance work.","Live use requires security controls and data-processing agreements appropriate to the actual jurisdiction and participants.","High-acuity or emergency dispatch remains out of scope.","No commercial platform license is assumed."],"source_ids":["S1","S3","S4","S6","S7"]},"annual_recurring":{"band_2026_usd":"50K_TO_250K","scope":"Hosting, monitoring, vulnerability management, adapter maintenance, conformance-suite updates, member support, incident response, governance and deprecation processes, and periodic privacy/leakage audits for a small federation.","confidence":"LOW","assumptions":["One to two shared full-time-equivalent roles plus infrastructure and specialist reviews.","Groups continue maintaining their own internal tools.","Traffic is modest and excludes regulated clinical records and emergency dispatch.","Volunteer labor is valued at replacement cost rather than treated as free."],"source_ids":["S3","S4","S5"]}},"verified_pipeline_gates":{"externally_supported_problem":{"status":"YES","reason":"Independent primary research and organizer-directed privacy guidance establish heterogeneous coordination tools, human dependence, privacy risk, and sociotechnical tension, though the exact status-error prevalence remains unknown.","source_ids":["S1","S2","S3","S7"]},"externally_credible_adopter_or_authorizer":{"status":"YES","reason":"mutua is a named, fiscally sponsored mutual-aid infrastructure initiative explicitly seeking community-controlled sharing protocols and inviting organizers to shape them. This establishes a credible prospective adopter, not an adoption commitment.","source_ids":["S4"]},"distinct_testable_incremental_claim":{"status":"YES","reason":"The remaining claim is contrastive against centralized closed-loop systems and directory standards and can be tested using independently built adapters, held-out histories, divergence counts, hidden-data-access logs, leakage checks, and member authority review.","source_ids":["S5","S6","S7","S8"]},"bounded_next_evidence_step":{"status":"YES","reason":"A two-week fictional-data tabletop with specified participants, histories, two independent adapters, comparators, recorded outcomes, stop conditions, and explicit falsifiers is bounded and reversible.","source_ids":["S3","S4","S8"]},"no_unresolved_safety_or_authority_stop":{"status":"YES","reason":"The proposed first step uses only fictional requests and volunteers, requires member-authorized representatives, preserves requester withdrawal, bars live routing and performance inference, and can revert to existing channels. Live deployment would require a new safety and legal review.","source_ids":["S1","S2","S3"]},"credible_cost_scope_and_range":{"status":"UNCERTAIN","reason":"The scopes and resource-equivalent bands are explicit and plausible relative to the demonstrated systems, but no direct labor quotes, integration inventory, jurisdiction, security classification, or adopter-specific capacity data were available.","source_ids":["S4","S5","S6","S7","S8"]}},"next_evidence_step":"Recruit authorized representatives from three real member groups plus one requester advocate and one federation dispatcher for a two-week, non-live tabletop. Freeze before implementation: the state vocabulary, authority matrix, public fields, notification side effects, privacy allowlist, acceptance thresholds, and stop rules. Independently implement a spreadsheet-backed adapter and a chat-task-backed adapter, then run 16 disclosed fictional histories plus at least 100 generated held-out transition sequences through both. Primary comparator: public state, error class, audit record, and notification outcome must match for every held-out sequence. Secondary comparators: current manual translation and a simple centralized shared spreadsheet, measured by dispatcher disagreement, duplicate active commitments, time to resolve state, requests for hidden data, and protected-field leakage. Falsify the intervention if conforming adapters diverge on any safety-critical transition; more than 1% of held-out sequences diverge overall; any routine history requires roster/chat inspection; protected volunteer or requester data appear in public outputs; or two or more authorized member/requester reviewers judge that the model converts negotiation into assignment or blocks withdrawal. Do not route live aid.","blocking_evidence":["No measured prevalence or baseline rate for cross-group status misclassification, duplicate dispatch, stale release, or hidden-record inspection in an actual federation.","No federation or three member groups have committed to participate or delegated semantic-change authority.","No evidence yet that legitimate negotiated, collective, or multi-provider commitments map to one abstract state without loss.","No independently implemented adapters have been tested against held-out transition sequences.","No adopter-specific tool/API inventory, identity model, notification reliability requirement, accessibility assessment, threat model, or jurisdictional privacy analysis exists.","No labor quotes or integration estimates validate the 2026 resource-equivalent cost bands.","No evidence yet shows advantage over a centralized shared spreadsheet, Unite Us/CAN-style platform, or continued human mediation."],"research_disposition":"PARTNERED_RESEARCH_PROGRAM","world_novelty_boundary":"The bounded search found substantial collision with closed-loop social-care referral platforms, coordinated-assistance workflows, Open Referral interoperability standards, and open-source mutual-aid applications. It did not locate a directly documented open contract combining provider-controlled acceptance/release authority, independently operated local adapters, representation-hiding privacy rules, and one shared behavioral conformance oracle for federated mutual-aid commitments. This is only a remaining contrast after eight-source web review: world novelty, patentability, freedom to operate, market size, and realized impact were not measured.","arm":"COMPLETE_PROPOSAL_PORTFOLIO","candidate_version":0,"controller_recommendation":{"action":"STOP_EMPIRICAL_RESEARCH_NEEDED","repairable":false,"material_progress_observed":true,"progress_targets":["Secure written participation and authority boundaries from three member groups, a requester advocate, and a federation dispatcher.","Measure manual-baseline disagreement, duplicate-active-commitment, stale-release, and hidden-data-inspection rates on fictional reconstructions of actual workflow patterns.","Demonstrate two independently implemented adapters with zero safety-critical divergence and no more than 1% overall divergence on at least 100 held-out sequences.","Show that routine routing needs no access to rosters, chat histories, private deliberation, or tool-specific fields and produces no protected-field leakage.","Obtain explicit member/requester judgments that acceptance remains voluntary, release remains provider-controlled, withdrawal remains available, and negotiated or multi-provider states are not erased.","Compare the contract against a centralized shared spreadsheet and human mediation on accuracy, coordination time, privacy exposure, and governance burden.","Produce an adopter-specific integration inventory, threat model, accessibility review, legal/privacy assessment, and externally checked cost estimate before any live pilot."],"reason":"Web evidence supports the broad problem, identifies a credible prospective adopter, and shows that implementation is technically plausible, but it also reveals substantial prior-art collision. The proposal's remaining value depends on empirical facts unavailable from further bounded web research: whether real groups can map their relational workflows without loss, whether independent adapters agree, whether privacy leakage is prevented, and whether member governance accepts the authority model. Those questions require partnered tabletop fieldwork and live-system knowledge, so the research stage must stop for empirical evidence."},"proposal_index":3}