{"schemaVersion":"1","name":"Handover research and evidence index","updatedAt":"2026-08-05","canonical":"https://handover.sh/research","humanView":"https://handover.sh/research","repositoryPackage":"https://github.com/44-pixels/handover-mcp/tree/main/research/v1","immutableRelease":"https://github.com/44-pixels/handover-mcp/releases/tag/research-v1.0.0","citation":{"cff":"https://github.com/44-pixels/handover-mcp/releases/download/research-v1.0.0/CITATION.cff","bibtex":"https://github.com/44-pixels/handover-mcp/releases/download/research-v1.0.0/citation.bib"},"independentEvidence":{"status":"open-no-accepted-results","acceptedResults":0,"benchmarkRelease":"https://github.com/44-pixels/handover-mcp/releases/tag/benchmark-v1.0.0","contributionGuide":"https://github.com/44-pixels/handover-mcp/tree/main/benchmark/v1/independent-results","issueForm":"https://github.com/44-pixels/handover-mcp/issues/new?template=independent-benchmark-result.yml","validator":"https://github.com/44-pixels/handover-mcp/blob/main/benchmark/v1/independent-results/validate.mjs","requirements":["Use one fresh session per prompt with tools, browsing, memory, and retrieval disabled.","Publish every raw response, including invalid output, failed parses, and declared retries.","Publish the original submission and deterministic scored result.","Declare material relationships and at least two limitations.","Exclude credentials, private context, customer data, and proprietary material."],"verificationBoundary":"Package validation proves completeness and deterministic score reproducibility. It does not prove model provenance or experimental compliance."},"publicationPolicy":{"inspectable":"Claims link to methods and source or structured artifacts.","bounded":"Material limitations and first-party relationships are stated beside results.","portable":"Evidence is published in directly reusable formats where possible."},"items":[{"slug":"ai-handoff-continuity-benchmark","label":"Open benchmark","title":"AI Handoff Continuity Benchmark","description":"A reproducible test of how much continuation-critical state a model recovers from a transcript, compressed memory, or structured handoff.","finding":"In a two-system authored pilot, structured handoffs scored 79.45, compared with 76.67 for conversation transcripts and 45.00 for compressed memory.","status":"Open pilot","measuredAt":"2026-08-04","href":"https://handover.sh/benchmark","image":"https://handover.sh/og/ai-handoff-benchmark-results.png","imageAlt":"AI Handoff Continuity Benchmark results comparing structured handoffs, transcripts, and compressed memory","schemaType":"Dataset","metrics":[{"value":"79.45","label":"structured handoff"},{"value":"76.67","label":"transcript"},{"value":"45.00","label":"compressed memory"}],"artifacts":[{"label":"Dataset","href":"https://handover.sh/benchmark/v1/dataset.json","format":"JSON"},{"label":"Result table","href":"https://handover.sh/benchmark/v1/results/2026-08-04/summary.csv","format":"CSV"},{"label":"Reproducibility record","href":"https://handover.sh/benchmark/v1/results/2026-08-04/README.md","format":"Markdown"}],"limitations":["Three authored cases and one run per model system do not estimate variance.","Candidate selection measures state recovery rather than free-form task quality.","Handover created the benchmark and benefits if structured handoffs perform well."]},{"slug":"reporter-migration-field-report","label":"Production field report","title":"Reporter to Handover migration","description":"A first-party production record of inventorying Reporter, verifying mapped handovers, checking file digests, and holding ambiguous records for review.","finding":"44pixels inventoried 1,497 source reports and verified 1,370 Handover records. All 400 comparable file hashes matched; 127 reports remained blocked for review.","status":"Migration validating","measuredAt":"2026-08-05","href":"https://handover.sh/case-studies/44pixels-reporter-to-handover-migration","image":"https://handover.sh/og/shared-knowledge-base-ai-agents.png","imageAlt":"Shared company knowledge base for people and AI agents in Handover","schemaType":"TechArticle","metrics":[{"value":"1,497","label":"reports inventoried"},{"value":"1,370","label":"records verified"},{"value":"400 / 400","label":"matching hashes"}],"artifacts":[{"label":"Field report","href":"https://handover.sh/case-studies/44pixels-reporter-to-handover-migration","format":"HTML"},{"label":"Agent evidence","href":"https://handover.sh/case-studies.json","format":"JSON"},{"label":"Migration checklist","href":"https://handover.sh/examples/static-report-migration-checklist.md","format":"Markdown"}],"limitations":["This is a first-party case study about one internal production corpus.","The migration remains blocked while ambiguous and restricted records are reviewed.","Digest checks prove byte equality for comparable files, not semantic equivalence of every report."]},{"slug":"handoff-continuity-record","label":"Open protocol","title":"Handoff Continuity Record v1.0","description":"A platform-neutral JSON record for the objective, state, decisions, evidence, constraints, next action, ownership, and open review needed to continue work.","finding":"The stable v1.0 format separates continuation state from routing, transport, authentication, storage, and tool invocation.","status":"Stable","measuredAt":"2026-08-03","href":"https://handover.sh/protocol","image":"https://handover.sh/og/ai-agent-handoff-protocol.png","imageAlt":"Open Handoff Continuity Record protocol for durable AI agent context","schemaType":"DigitalDocument","metrics":[{"value":"1.0","label":"stable version"},{"value":"8","label":"continuity sections"},{"value":"MIT","label":"license"}],"artifacts":[{"label":"JSON Schema","href":"https://handover.sh/protocol/v1/schema.json","format":"JSON"},{"label":"Valid example","href":"https://handover.sh/protocol/v1/example.json","format":"JSON"},{"label":"Validator","href":"https://handover.sh/protocol/v1/validate.mjs","format":"JavaScript"}],"limitations":["The protocol does not define transport, routing, authentication, or storage.","Structural conformance does not prove the evidence is true or complete.","Receivers must still verify access, freshness, and acceptance criteria."]},{"slug":"public-continuation-demo","label":"Runnable evidence","title":"Human-agent continuation demo","description":"A public, no-login example showing multiple actors carrying a reporting workflow through immutable revisions, files, anchored review, and a verified next action.","finding":"The demo exposes the same continuation record to a human reader and a receiving agent through HTML, JSON, Markdown, SQL, and a bounded successor prompt.","status":"Public","measuredAt":"2026-08-03","href":"https://handover.sh/demo","image":"https://handover.sh/screenshots/demo-review-and-revisions.png","imageAlt":"Handover demo showing immutable revisions and an anchored review thread","schemaType":"SoftwareApplication","metrics":[{"value":"3","label":"actors"},{"value":"3","label":"revisions"},{"value":"1","label":"resolved review"}],"artifacts":[{"label":"Demo manifest","href":"https://handover.sh/demo.json","format":"JSON"},{"label":"Context","href":"https://handover.sh/demo/context.md","format":"Markdown"},{"label":"Inventory query","href":"https://handover.sh/demo/inventory.sql","format":"SQL"},{"label":"Successor prompt","href":"https://handover.sh/demo/prompt.md","format":"Markdown"}],"limitations":["The demo is a curated public workflow rather than an independent user study.","It demonstrates inspectability and continuation mechanics, not task-quality improvement.","Production access policy and private organization data are intentionally excluded."]}]}