Working patterns
Start with a complete workflow, benchmark, review pattern, or portable context record. Filter by the host and artifact already in your work.
Agent handoff
Open a no-login demo containing Markdown, SQL, JSON, a visual artifact, three attributable revisions, and a resolved review thread.
Every artifact is downloadable, and the same workflow is available as an agent-readable JSON record.
Agent handoff
Publish a structured handoff, retrieve it from another host, and append the next revision without losing decisions or ownership.
A complete, copyable workflow with revision checks and an explicit continuation step.
Agent handoff
Publish Markdown, SQL, and JSON, review the exact revision, correct the evidence with optimistic concurrency, resolve the finding, and prove a successor can continue.
The raw recipe uses the production Handover tool schemas and includes identity, artifact, review, correction, resolution, denied-access, stale-revision, and pass-criteria checks.
Context portability
Configure Claude Code, Cursor, and Codex with separate named credentials, then prove a second host can retrieve and continue the first host's work.
The raw recipe includes copyable host configuration, identity checks, an end-to-end continuity test, pass criteria, and a security checklist.
Connection verification
Configure Cline IDE or CLI with remote Streamable HTTP, verify a named service identity, and prove a fresh successor can retrieve and continue the exact artifact.
The raw workflow covers Cline's wizard and private IDE configuration, protected reads, an approved write and read-back, two-identity continuation, denied access, stale revisions, and revocation.
Connection verification
Add a user-scoped remote HTTP server, diagnose disabled and disconnected states, verify a named service identity, and prove a fresh successor can continue the exact artifact.
The raw workflow covers the tested Gemini CLI command and saved configuration, protected reads, an approved write and read-back, two-identity continuation, denied access, stale revisions, revocation, and complete local plus remote removal.
Connection verification
Add a user-scoped remote HTTP server, diagnose authentication, trust, and enterprise-policy states, then prove a fresh successor can continue the exact artifact.
The raw workflow covers the tested Claude Code command and saved configuration, protected reads, approved write and read-back, two-identity continuation, denied access, stale revisions, credential exposure boundaries, and complete removal.
Connection verification
Configure a remote Streamable HTTP server for Cursor IDE and Agent CLI, diagnose connection, environment, disabled, and OAuth states, then prove a fresh successor can continue the exact artifact.
The raw workflow covers global and project configuration, a credential-free one-click installer, current Agent CLI diagnostics, protected reads, approved write and read-back, two-identity continuation, disabled and denied states, stale revisions, and complete removal.
Connection verification
Add a remote Streamable HTTP server to Codex CLI, the IDE extension, or the desktop app, verify a named service identity, and prove a fresh successor can continue the exact artifact.
The raw workflow covers the tested Codex CLI lifecycle, environment-backed credentials, global and trusted-project scope, protected reads, approved write and read-back, two-identity continuation, denied and stale states, ChatGPT web boundaries, revocation, and complete removal.
Context portability
Separate durable repository instructions from changing task state, checkpoint Git evidence, and give the successor an exact verification sequence.
The guide and downloadable workflow cover AGENTS.md, CLAUDE.md, GEMINI.md, Cursor rules, HANDOFF.md, Git state, and Handover.
Context portability
Choose between Codex's one-time Claude Code import and a repeatable, revisioned handoff that a fresh Codex session can verify and continue.
The raw recipe includes separate identities, Git and instruction-file checks, artifact read-back, optimistic concurrency, revocation, and explicit pass criteria.
Context portability
Install the Handover command, authenticate once, publish a folder, and pull the same context into another working session.
The public npm package supports Markdown, HTML, SQL, images, and binary files.
Evaluation
Compare transcript, compressed-memory, and structured-handoff conditions with a deterministic continuation benchmark.
The open pilot includes a versioned dataset, answer key, runner, and scoring method.
Human review
Publish an artifact, anchor actionable feedback, mention or assign an authorized resolver, and link every resolved finding to the corrected revision.
The raw workflow includes exact MCP operations, an annotation payload, concurrency checks, resolution metadata, and pass criteria.
Identity governance
Document every service agent, credential owner, scope, rotation date, and revocation path in one reusable record.
The template separates human attribution from named, revocable agent credentials.
Identity governance
Assign an owner and risk tier, map identity and access, define approval gates, retain evidence, test controls, and revoke the agent cleanly.
The raw checklist includes a copyable access matrix, tool controls, approval record, continuation evidence, authorization tests, incident response, and offboarding gates.
Import and export
Export a complete, versioned context bundle and import it into another tool without making Handover the only readable copy.
The workflow preserves files and metadata through an interoperable archive.
Import and export
Inventory products, folders, HTML entry points, nested assets, owners, audiences, URLs, and publishing clients before a reversible dual-run cutover.
The checklist includes source manifests, mapping rules, hash and rendering checks, human-agent acceptance, and a rollback gate.
Company operations
Transfer responsibility between teams, people, and AI agents with explicit scope, access classification, evidence, receiver acceptance, continuation, and rollback.
The copyable runbook and machine-readable control record require independent receiver authentication, exact artifact read-back, one reproduced material claim, denied and revoked access checks, and a fresh successor revision.
Context engineering
Decide whether information belongs in instructions, active working context, runtime memory, shared context, or source artifacts.
The open Markdown and JSON records include placement fields, decision rules, security boundaries, retention controls, and a two-identity continuity test.
Agent Skills
Package one reliable Handover workflow in SKILL.md, connect it to authenticated MCP capabilities, and test identity, success, denied access, continuity, and revocation.
The raw implementation asset includes a folder contract, frontmatter, tool sequence, boundary rules, two-host test, publication checklist, and install commands.
Agent Skills
Install a portable SKILL.md, verify the publishing identity, preserve unfinished work, and prove a separately authenticated successor can retrieve and continue it.
The raw workflow covers natural-language activation, exact artifact read-back, cross-host continuation, denied and read-only access, stale revisions, and credential revocation.
Agent Skills
Install the same Handover skill into Claude Code, Codex, Cursor, or Gemini CLI, then verify discovery, activation, MCP identity, read-back, denied access, and cross-host continuation.
The raw checklist includes tested host names and project paths, native discovery checks, an authenticated trigger test, pass criteria, troubleshooting, and revocation.
Connection verification
Check the endpoint, authenticated principal, workspace, role, scopes, and a protected context read, then run a two-identity continuation test.
The raw checklist includes human-readable and JSON diagnostics, failure classification, a write boundary, and end-to-end pass criteria.