You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
Repository navigation
[upstream-adapt] Receive Drive push notifications with durable cursors and hooks #204
Operators can run a secure Drive push-notification receiver with persisted cursors, authenticated callbacks, and optional channel renewal for near-real-time automation.
Why
Upstream drive watch receiver is advanced automation infrastructure; only if we need live Drive hooks beyond Gmail watch.
Context
Source: upstream feature review of openclaw/gogcli (2026-08-11)
Intake rule: adapt the outcome in our architecture — do not cherry-pick or copy upstream implementation
Suggested triage priority: P3
Area: Drive
Non-goals for this issue
No prescribed flags, package layout, or upstream file ports
No requirement to match upstream UX naming
Implementation design happens during triage / agent brief
Acceptance (triage will refine)
Desired user/agent-visible outcome is clear and still wanted
Not already fully covered by current main
Scope is bounded enough for an agent brief (or split)
Drive exposes change listing and provider watch/channel setup primitives, while Gmail has a service-specific receiver and durable watch state. There is no Drive callback receiver that validates notifications, advances persisted change cursors, dispatches hooks, or renews channels. No local historical request implements that combined outcome.
The candidate is advanced infrastructure spanning public callback hosting, authentication, change-page replay, cursor durability, duplicate delivery, hook secret handling, channel expiration/renewal, shared-drive coverage, and crash recovery. Its deployment and trust assumptions are unspecified and materially affect the design.
Recommendation: retain needs-triage until the maintainer confirms a live Drive-hook use case and defines receiver deployment, callback authentication, cursor ownership, delivery guarantees, and renewal scope. Split receiver, durable change processing, outbound hooks, and renewal if accepted.
Drive already exposes start-page tokens, paged change listing, provider webhook-channel creation for changes and files, and channel stopping. Gmail separately has a callback server and durable watch state. The snapshot does not provide a Drive receiver that authenticates callbacks, advances a persisted change cursor, replays pages, dispatches hooks, or renews channels. Historical records contain no implementation of that combined outcome or prior rejection.
The missing surface spans public callback deployment, notification authentication, cursor ownership and atomicity, duplicate and out-of-order delivery, crash recovery, hook-secret handling, shared-drive coverage, delivery guarantees, and expiration renewal. These are deployment and trust decisions rather than incidental implementation details.
Recommendation: retain needs-triage until the live Drive-hook use case, receiver deployment model, callback authentication, cursor contract, delivery guarantees, and renewal scope are approved; split receiver, change processing, outbound hooks, and renewal if accepted.
Drive already exposes start-page tokens, paged change listing, webhook-channel creation for Drive changes and files, and channel stopping. Gmail separately has an authenticated callback server, atomic durable watch state, replay handling, duplicate suppression, and optional outbound hooks. The snapshot does not provide a Drive receiver that authenticates callbacks, advances a persisted change cursor, replays pages, dispatches hooks, or renews channels.
Historical records show those primitives but no implementation of the combined Drive outcome, and no prior-rejection record exists.
The missing surface spans public callback deployment, notification authentication, cursor ownership and atomicity, duplicate and out-of-order delivery, crash recovery, hook-secret handling, shared-drive coverage, delivery guarantees, and channel expiration and renewal. These are deployment and trust decisions rather than incidental implementation details.
Recommendation: retain needs-triage until the live Drive-hook use case, receiver deployment model, callback authentication, cursor contract, delivery guarantees, and renewal scope are approved; split receiver, change processing, outbound hooks, and renewal if accepted.
Drive exposes start-page tokens, paged change listing, webhook-channel creation for changes and files, and channel stopping. Gmail separately has an authenticated callback server, atomic durable watch state, replay handling, duplicate suppression, and optional outbound hooks. The snapshot has no Drive receiver that authenticates callbacks, advances a persisted cursor, replays pages, dispatches hooks, or renews channels. Historical records show the primitives but no duplicate or prior rejection.
The missing surface spans public callback deployment, notification authentication, cursor ownership and atomicity, duplicate and out-of-order delivery, crash recovery, hook-secret handling, shared-drive coverage, delivery guarantees, and channel expiration and renewal. These are deployment and trust decisions.
Recommendation: retain needs-triage until the live Drive-hook use case, receiver deployment model, callback authentication, cursor contract, delivery guarantees, and renewal scope are approved; split receiver, change processing, outbound hooks, and renewal if accepted.
Drive exposes start-page tokens, paged change listing, webhook-channel creation for changes and files, and channel stopping. Gmail separately has an authenticated callback server, atomic durable watch state, replay handling, duplicate suppression, and optional outbound hooks. The snapshot has no Drive receiver that authenticates callbacks, advances a persisted cursor, replays pages, dispatches hooks, or renews channels. Historical records show the primitives but no duplicate or prior rejection.
The missing surface spans public callback deployment, notification authentication, cursor ownership and atomicity, duplicate and out-of-order delivery, crash recovery, hook-secret handling, shared-drive coverage, delivery guarantees, and channel expiration and renewal. These are deployment and trust decisions. Retain needs-triage until the live Drive-hook use case, receiver deployment model, callback authentication, cursor contract, delivery guarantees, and renewal scope are approved; split receiver, change processing, outbound hooks, and renewal if accepted.
Drive exposes start-page tokens, paged change listing, webhook-channel creation for changes and files, and channel stopping. Gmail separately has an authenticated callback server, atomic durable watch state, replay handling, duplicate suppression, and optional outbound hooks. The snapshot has no Drive receiver that authenticates callbacks, advances a persisted cursor, replays pages, dispatches hooks, or renews channels. Historical records show the primitives but no duplicate or prior-rejection record.
The missing surface spans public callback deployment, notification authentication, cursor ownership and atomicity, duplicate and out-of-order delivery, crash recovery, hook-secret handling, shared-drive coverage, delivery guarantees, and channel expiration and renewal. These are deployment and trust decisions. Retain needs-triage until the live Drive-hook use case, receiver deployment model, callback authentication, cursor contract, delivery guarantees, and renewal scope are approved; split receiver, change processing, outbound hooks, and renewal if accepted.
Drive exposes start-page tokens, paged change listing, webhook-channel creation for changes and files, and channel stopping. Gmail separately has an authenticated callback server, atomic durable watch state, replay handling, duplicate suppression, and optional outbound hooks. The snapshot has no Drive receiver that authenticates callbacks, advances a persisted cursor, replays pages, dispatches hooks, or renews channels. The supplied historical snapshots show the primitives but no duplicate implementation, and there is no prior-rejection record.
The missing surface spans public callback deployment, notification authentication, cursor ownership and atomicity, duplicate and out-of-order delivery, crash recovery, hook-secret handling, shared-drive coverage, delivery guarantees, and channel expiration and renewal. These are deployment and trust decisions. Retain needs-triage until the live Drive-hook use case, receiver deployment model, callback authentication, cursor contract, delivery guarantees, and renewal scope are approved; split receiver, change processing, outbound hooks, and renewal if accepted.
Drive exposes start-page tokens, paged change listing, webhook-channel creation for changes and files, and channel stopping. Gmail separately has an authenticated callback server and durable watch state. The snapshot has no Drive receiver that authenticates callbacks, advances a persisted cursor, replays pages, dispatches hooks, or renews channels. The historical snapshots show those primitives but no duplicate implementation, and there is no prior-rejection record.
The missing surface spans public callback deployment, notification authentication, cursor ownership and atomicity, duplicate and out-of-order delivery, crash recovery, hook-secret handling, shared-drive coverage, delivery guarantees, and channel expiration and renewal. These are deployment and trust decisions. Retain needs-triage until the live Drive-hook use case, receiver deployment model, callback authentication, cursor contract, delivery guarantees, and renewal scope are approved; split receiver, change processing, outbound hooks, and renewal if accepted.
Drive exposes start-page tokens, paged change listing, webhook-channel creation for changes and files, and channel stopping. Gmail separately has an authenticated callback server and durable watch state. The current snapshot has no Drive receiver that authenticates callbacks, advances a persisted cursor, replays pages, dispatches hooks, or renews channels. The supplied histories show those primitives but no duplicate implementation, and there is no prior-rejection record.
The missing surface spans public callback deployment, notification authentication, cursor ownership and atomicity, duplicate and out-of-order delivery, crash recovery, hook-secret handling, shared-drive coverage, delivery guarantees, and channel expiration and renewal. These are deployment and trust decisions. Retain needs-triage until the live Drive-hook use case, receiver deployment model, callback authentication, cursor contract, delivery guarantees, and renewal scope are approved; split receiver, change processing, outbound hooks, and renewal if accepted.
Drive exposes start-page tokens, paged change listing, webhook-channel creation for changes and files, and channel stopping. Gmail separately has an authenticated callback server and durable watch state. The source has no Drive receiver that authenticates callbacks, advances a persisted cursor, replays pages, dispatches hooks, or renews channels. Searches of the supplied histories found those primitives but no duplicate implementation; no prior-rejection record exists.
The missing surface spans public callback deployment, notification authentication, cursor ownership and atomicity, duplicate and out-of-order delivery, crash recovery, hook-secret handling, shared-drive coverage, delivery guarantees, and channel expiration and renewal. These are deployment and trust decisions. Retain needs-triage until the live Drive-hook use case, receiver deployment model, callback authentication, cursor contract, delivery guarantees, and renewal scope are approved; split receiver, change processing, outbound hooks, and renewal if accepted.
Drive exposes start-page tokens, paged change listing, webhook-channel creation, and channel stopping. Gmail separately has an authenticated callback server and durable watch state. There is no Drive receiver that authenticates callbacks, advances a persisted cursor, replays pages, dispatches hooks, or renews channels. Searches of the supplied histories found those primitives but no duplicate implementation, and there is no prior-rejection record.
The missing surface spans callback deployment, notification authentication, cursor ownership and atomicity, duplicate and out-of-order delivery, crash recovery, hook-secret handling, shared-drive coverage, delivery guarantees, and renewal. Retain needs-triage until the live Drive-hook use case, deployment model, callback authentication, cursor contract, delivery guarantees, and renewal scope are approved; split receiver, change processing, hooks, and renewal if accepted.
Drive exposes start-page tokens, paged change listing, webhook-channel creation, and channel stopping. Gmail separately has an authenticated callback server and durable watch state. There is no Drive receiver that authenticates callbacks, advances a persisted cursor, replays pages, dispatches hooks, or renews channels. Searches of Drive changes/channels, Gmail watch, persistence, and hook concepts and the supplied closed-issue and merged-PR histories found those primitives but no duplicate implementation; no prior-rejection record exists.
The missing surface spans callback deployment, notification authentication, cursor ownership and atomicity, duplicate and out-of-order delivery, crash recovery, hook-secret handling, shared-drive coverage, delivery guarantees, and renewal. A maintainer must approve the live Drive-hook use case, deployment model, callback authentication, cursor contract, delivery guarantees, and renewal scope; receiver, change processing, hooks, and renewal should be split if accepted. Keep this in needs-triage until those decisions are recorded.
Drive exposes start-page tokens, paged change listing, webhook-channel creation, and channel stopping. Gmail separately has an authenticated callback server and durable watch state. There is no Drive receiver that authenticates callbacks, advances a persisted cursor, replays pages, dispatches hooks, or renews channels. Concept searches of Drive changes and channels, Gmail watch, persistence, hooks, and the supplied histories found those primitives but no duplicate implementation, and there is no prior-rejection record.
The missing surface spans callback deployment, notification authentication, cursor ownership and atomicity, duplicate and out-of-order delivery, crash recovery, hook-secret handling, shared-drive coverage, delivery guarantees, and renewal. A maintainer must approve the live Drive-hook use case, deployment model, callback authentication, cursor contract, delivery guarantees, and renewal scope; receiver, change processing, hooks, and renewal should be split if accepted. Retain needs-triage until those decisions are recorded.
Category: enhancement Summary: Operators can run a secure Drive push-notification receiver with persisted cursors, authenticated callbacks, and optional channel renewal for near-real-time automation.
Current behavior:
Fork lacks this upstream-adapt capability (or only partial primitives exist). Prior triage notes on this issue document the current surface.
Desired behavior:
Operators can run a secure Drive push-notification receiver with persisted cursors, authenticated callbacks, and optional channel renewal for near-real-time automation.
Key interfaces:
Existing gog CLI command surface and auth/safety controls
Google API surface named in the issue outcome
Script-friendly machine-readable output consistent with sibling commands
Acceptance criteria:
Outcome in the issue body is met under this fork’s architecture (adapt, do not cherry-pick upstream patches)
Behavior is covered by tests or documented manual verification
Safety/auth constraints match the rest of gog
Out of scope:
Copying upstream implementation wholesale
Unrelated Google services
Why not delegable
Product priority and contract semantics are still open (what ships first, audit/lifecycle meaning, and whether this fork needs the capability at all). An agent should not invent those decisions. Once Jeremy ranks or rejects this upstream-adapt item, retarget to ready-for-agent or wontfix.
Prior triage notes remain the technical baseline; this comment only advances state out of needs-triage.
Closing as obsolete (GitHub sweep 2026-10-12). This [upstream-adapt] request asked to port an upstream steipete/gogcli CLI feature into our gog CLI. That CLI was retired in 91336b7 (#269); main builds only gog-mcp, so the command surface this request targets no longer exists. New needs on the MCP side are filed as MCP-scoped issues (see #271, #273, #274, #278).
Activity
itsjeremyjohnson commentedon Aug 12, 2026
Triage Notes
Drive exposes change listing and provider watch/channel setup primitives, while Gmail has a service-specific receiver and durable watch state. There is no Drive callback receiver that validates notifications, advances persisted change cursors, dispatches hooks, or renews channels. No local historical request implements that combined outcome.
The candidate is advanced infrastructure spanning public callback hosting, authentication, change-page replay, cursor durability, duplicate delivery, hook secret handling, channel expiration/renewal, shared-drive coverage, and crash recovery. Its deployment and trust assumptions are unspecified and materially affect the design.
Recommendation: retain
needs-triageuntil the maintainer confirms a live Drive-hook use case and defines receiver deployment, callback authentication, cursor ownership, delivery guarantees, and renewal scope. Split receiver, durable change processing, outbound hooks, and renewal if accepted.itsjeremyjohnson commentedon Aug 13, 2026
Triage Notes
Drive already exposes start-page tokens, paged change listing, provider webhook-channel creation for changes and files, and channel stopping. Gmail separately has a callback server and durable watch state. The snapshot does not provide a Drive receiver that authenticates callbacks, advances a persisted change cursor, replays pages, dispatches hooks, or renews channels. Historical records contain no implementation of that combined outcome or prior rejection.
The missing surface spans public callback deployment, notification authentication, cursor ownership and atomicity, duplicate and out-of-order delivery, crash recovery, hook-secret handling, shared-drive coverage, delivery guarantees, and expiration renewal. These are deployment and trust decisions rather than incidental implementation details.
Recommendation: retain
needs-triageuntil the live Drive-hook use case, receiver deployment model, callback authentication, cursor contract, delivery guarantees, and renewal scope are approved; split receiver, change processing, outbound hooks, and renewal if accepted.itsjeremyjohnson commentedon Aug 14, 2026
Triage Notes
Drive already exposes start-page tokens, paged change listing, webhook-channel creation for Drive changes and files, and channel stopping. Gmail separately has an authenticated callback server, atomic durable watch state, replay handling, duplicate suppression, and optional outbound hooks. The snapshot does not provide a Drive receiver that authenticates callbacks, advances a persisted change cursor, replays pages, dispatches hooks, or renews channels.
Historical records show those primitives but no implementation of the combined Drive outcome, and no prior-rejection record exists.
The missing surface spans public callback deployment, notification authentication, cursor ownership and atomicity, duplicate and out-of-order delivery, crash recovery, hook-secret handling, shared-drive coverage, delivery guarantees, and channel expiration and renewal. These are deployment and trust decisions rather than incidental implementation details.
Recommendation: retain
needs-triageuntil the live Drive-hook use case, receiver deployment model, callback authentication, cursor contract, delivery guarantees, and renewal scope are approved; split receiver, change processing, outbound hooks, and renewal if accepted.itsjeremyjohnson commentedon Aug 15, 2026
Triage Notes
Drive exposes start-page tokens, paged change listing, webhook-channel creation for changes and files, and channel stopping. Gmail separately has an authenticated callback server, atomic durable watch state, replay handling, duplicate suppression, and optional outbound hooks. The snapshot has no Drive receiver that authenticates callbacks, advances a persisted cursor, replays pages, dispatches hooks, or renews channels. Historical records show the primitives but no duplicate or prior rejection.
The missing surface spans public callback deployment, notification authentication, cursor ownership and atomicity, duplicate and out-of-order delivery, crash recovery, hook-secret handling, shared-drive coverage, delivery guarantees, and channel expiration and renewal. These are deployment and trust decisions.
Recommendation: retain
needs-triageuntil the live Drive-hook use case, receiver deployment model, callback authentication, cursor contract, delivery guarantees, and renewal scope are approved; split receiver, change processing, outbound hooks, and renewal if accepted.itsjeremyjohnson commentedon Aug 16, 2026
Triage Notes
Drive exposes start-page tokens, paged change listing, webhook-channel creation for changes and files, and channel stopping. Gmail separately has an authenticated callback server, atomic durable watch state, replay handling, duplicate suppression, and optional outbound hooks. The snapshot has no Drive receiver that authenticates callbacks, advances a persisted cursor, replays pages, dispatches hooks, or renews channels. Historical records show the primitives but no duplicate or prior rejection.
The missing surface spans public callback deployment, notification authentication, cursor ownership and atomicity, duplicate and out-of-order delivery, crash recovery, hook-secret handling, shared-drive coverage, delivery guarantees, and channel expiration and renewal. These are deployment and trust decisions. Retain
needs-triageuntil the live Drive-hook use case, receiver deployment model, callback authentication, cursor contract, delivery guarantees, and renewal scope are approved; split receiver, change processing, outbound hooks, and renewal if accepted.itsjeremyjohnson commentedon Aug 16, 2026
Triage Notes
Drive exposes start-page tokens, paged change listing, webhook-channel creation for changes and files, and channel stopping. Gmail separately has an authenticated callback server, atomic durable watch state, replay handling, duplicate suppression, and optional outbound hooks. The snapshot has no Drive receiver that authenticates callbacks, advances a persisted cursor, replays pages, dispatches hooks, or renews channels. Historical records show the primitives but no duplicate or prior-rejection record.
The missing surface spans public callback deployment, notification authentication, cursor ownership and atomicity, duplicate and out-of-order delivery, crash recovery, hook-secret handling, shared-drive coverage, delivery guarantees, and channel expiration and renewal. These are deployment and trust decisions. Retain
needs-triageuntil the live Drive-hook use case, receiver deployment model, callback authentication, cursor contract, delivery guarantees, and renewal scope are approved; split receiver, change processing, outbound hooks, and renewal if accepted.itsjeremyjohnson commentedon Aug 17, 2026
Triage Notes
Drive exposes start-page tokens, paged change listing, webhook-channel creation for changes and files, and channel stopping. Gmail separately has an authenticated callback server, atomic durable watch state, replay handling, duplicate suppression, and optional outbound hooks. The snapshot has no Drive receiver that authenticates callbacks, advances a persisted cursor, replays pages, dispatches hooks, or renews channels. The supplied historical snapshots show the primitives but no duplicate implementation, and there is no prior-rejection record.
The missing surface spans public callback deployment, notification authentication, cursor ownership and atomicity, duplicate and out-of-order delivery, crash recovery, hook-secret handling, shared-drive coverage, delivery guarantees, and channel expiration and renewal. These are deployment and trust decisions. Retain
needs-triageuntil the live Drive-hook use case, receiver deployment model, callback authentication, cursor contract, delivery guarantees, and renewal scope are approved; split receiver, change processing, outbound hooks, and renewal if accepted.itsjeremyjohnson commentedon Aug 18, 2026
Triage Notes
Drive exposes start-page tokens, paged change listing, webhook-channel creation for changes and files, and channel stopping. Gmail separately has an authenticated callback server and durable watch state. The snapshot has no Drive receiver that authenticates callbacks, advances a persisted cursor, replays pages, dispatches hooks, or renews channels. The historical snapshots show those primitives but no duplicate implementation, and there is no prior-rejection record.
The missing surface spans public callback deployment, notification authentication, cursor ownership and atomicity, duplicate and out-of-order delivery, crash recovery, hook-secret handling, shared-drive coverage, delivery guarantees, and channel expiration and renewal. These are deployment and trust decisions. Retain
needs-triageuntil the live Drive-hook use case, receiver deployment model, callback authentication, cursor contract, delivery guarantees, and renewal scope are approved; split receiver, change processing, outbound hooks, and renewal if accepted.itsjeremyjohnson commentedon Aug 20, 2026
Triage Notes
Drive exposes start-page tokens, paged change listing, webhook-channel creation for changes and files, and channel stopping. Gmail separately has an authenticated callback server and durable watch state. The current snapshot has no Drive receiver that authenticates callbacks, advances a persisted cursor, replays pages, dispatches hooks, or renews channels. The supplied histories show those primitives but no duplicate implementation, and there is no prior-rejection record.
The missing surface spans public callback deployment, notification authentication, cursor ownership and atomicity, duplicate and out-of-order delivery, crash recovery, hook-secret handling, shared-drive coverage, delivery guarantees, and channel expiration and renewal. These are deployment and trust decisions. Retain
needs-triageuntil the live Drive-hook use case, receiver deployment model, callback authentication, cursor contract, delivery guarantees, and renewal scope are approved; split receiver, change processing, outbound hooks, and renewal if accepted.itsjeremyjohnson commentedon Aug 21, 2026
Triage Notes
Drive exposes start-page tokens, paged change listing, webhook-channel creation for changes and files, and channel stopping. Gmail separately has an authenticated callback server and durable watch state. The source has no Drive receiver that authenticates callbacks, advances a persisted cursor, replays pages, dispatches hooks, or renews channels. Searches of the supplied histories found those primitives but no duplicate implementation; no prior-rejection record exists.
The missing surface spans public callback deployment, notification authentication, cursor ownership and atomicity, duplicate and out-of-order delivery, crash recovery, hook-secret handling, shared-drive coverage, delivery guarantees, and channel expiration and renewal. These are deployment and trust decisions. Retain
needs-triageuntil the live Drive-hook use case, receiver deployment model, callback authentication, cursor contract, delivery guarantees, and renewal scope are approved; split receiver, change processing, outbound hooks, and renewal if accepted.itsjeremyjohnson commentedon Aug 23, 2026
Triage Notes
Drive exposes start-page tokens, paged change listing, webhook-channel creation, and channel stopping. Gmail separately has an authenticated callback server and durable watch state. There is no Drive receiver that authenticates callbacks, advances a persisted cursor, replays pages, dispatches hooks, or renews channels. Searches of the supplied histories found those primitives but no duplicate implementation, and there is no prior-rejection record.
The missing surface spans callback deployment, notification authentication, cursor ownership and atomicity, duplicate and out-of-order delivery, crash recovery, hook-secret handling, shared-drive coverage, delivery guarantees, and renewal. Retain
needs-triageuntil the live Drive-hook use case, deployment model, callback authentication, cursor contract, delivery guarantees, and renewal scope are approved; split receiver, change processing, hooks, and renewal if accepted.itsjeremyjohnson commentedon Aug 24, 2026
Triage Notes
Drive exposes start-page tokens, paged change listing, webhook-channel creation, and channel stopping. Gmail separately has an authenticated callback server and durable watch state. There is no Drive receiver that authenticates callbacks, advances a persisted cursor, replays pages, dispatches hooks, or renews channels. Searches of Drive changes/channels, Gmail watch, persistence, and hook concepts and the supplied closed-issue and merged-PR histories found those primitives but no duplicate implementation; no prior-rejection record exists.
The missing surface spans callback deployment, notification authentication, cursor ownership and atomicity, duplicate and out-of-order delivery, crash recovery, hook-secret handling, shared-drive coverage, delivery guarantees, and renewal. A maintainer must approve the live Drive-hook use case, deployment model, callback authentication, cursor contract, delivery guarantees, and renewal scope; receiver, change processing, hooks, and renewal should be split if accepted. Keep this in
needs-triageuntil those decisions are recorded.itsjeremyjohnson commentedon Aug 25, 2026
Triage Notes
Drive exposes start-page tokens, paged change listing, webhook-channel creation, and channel stopping. Gmail separately has an authenticated callback server and durable watch state. There is no Drive receiver that authenticates callbacks, advances a persisted cursor, replays pages, dispatches hooks, or renews channels. Concept searches of Drive changes and channels, Gmail watch, persistence, hooks, and the supplied histories found those primitives but no duplicate implementation, and there is no prior-rejection record.
The missing surface spans callback deployment, notification authentication, cursor ownership and atomicity, duplicate and out-of-order delivery, crash recovery, hook-secret handling, shared-drive coverage, delivery guarantees, and renewal. A maintainer must approve the live Drive-hook use case, deployment model, callback authentication, cursor contract, delivery guarantees, and renewal scope; receiver, change processing, hooks, and renewal should be split if accepted. Retain
needs-triageuntil those decisions are recorded.itsjeremyjohnson commentedon Sep 15, 2026
Agent Brief
Category: enhancement
Summary: Operators can run a secure Drive push-notification receiver with persisted cursors, authenticated callbacks, and optional channel renewal for near-real-time automation.
Current behavior:
Fork lacks this upstream-adapt capability (or only partial primitives exist). Prior triage notes on this issue document the current surface.
Desired behavior:
Operators can run a secure Drive push-notification receiver with persisted cursors, authenticated callbacks, and optional channel renewal for near-real-time automation.
Key interfaces:
Acceptance criteria:
Out of scope:
Why not delegable
Product priority and contract semantics are still open (what ships first, audit/lifecycle meaning, and whether this fork needs the capability at all). An agent should not invent those decisions. Once Jeremy ranks or rejects this upstream-adapt item, retarget to
ready-for-agentorwontfix.Prior triage notes remain the technical baseline; this comment only advances state out of
needs-triage.itsjeremyjohnson commentedon Oct 12, 2026
Closing as obsolete (GitHub sweep 2026-10-12). This [upstream-adapt] request asked to port an upstream steipete/gogcli CLI feature into our
gogCLI. That CLI was retired in 91336b7 (#269);mainbuilds onlygog-mcp, so the command surface this request targets no longer exists. New needs on the MCP side are filed as MCP-scoped issues (see #271, #273, #274, #278).