roadmap: Phase 2 → Value Loops (user-value-first evolution) - #52
Conversation
Reorders Phase 2 around complete vertical loops (templates → meeting minutes → report-to-deck) instead of horizontal agent infrastructure: - Live meeting minutes promoted to keystone demo of the phase - Report → Deck (cross-generation) pulled forward from Phase 4 - New: Document health check (light proactive agent) + Deliverable pack - MCP server stays P0 but becomes the spine built in parallel - Deferred: role agents, runtime plugins, artifact generation (subsumed) - Metrics: time raw-material→deliverable <30min, loops/week - Fix upstream link: genspark-ai/genoffice (was hermesoffice)
|
Bugbot is not enabled for your account, so this pull request was not reviewed. Enable Bugbot in the Cursor dashboard to get automatic reviews on future PRs. |
![chatgpt-codex-connector[bot]](https://avatars.githubusercontent.com/in/1144995?s=60&v=4){kind=small}
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: 6978d35790
ℹ️ About Codex in GitHub
Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you
- Open a pull request for review
- Mark a draft as ready
- Comment "@codex review".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".
| **Sequencing:** the phase opens with the cheapest loop (templates, already in | ||
| flight), then the keystone demo — *a 45-minute meeting ends with a complete, | ||
| local minutes document, decisions and action items with owners, captured | ||
| without note-taking* — then Report → Deck. The MCP server builds in parallel |
There was a problem hiding this comment.
This newly makes live 45-minute capture/STT the keystone immediately after templates, but docs/product/agent-native-priorities.md:44-52 still directs contributors to start with transcript import and add live audio/STT only after the proposal flow is stable. Contributors following that implementation plan will take the opposite critical path, so update it or explicitly mark it superseded as part of this roadmap reorder.
Useful? React with 👍 / 👎.
| | 1 | download → first agent-assisted document < 10 min; CI green ≥95% of days; ≥5 external PRs merged | | ||
| | Trusted Agent Actions | share of agent mutations accepted vs reverted (target ≥70% after the first iteration); time from opening a document to the first agent action | | ||
| | 2 | keystone demo ships (external agent edits a deck through the trust pipeline); external MCP clients acting on real documents; artifacts generated per agent session | | ||
| | 2 | keystone demo ships (45-min meeting → complete local minutes docx); report → deck pipeline; time from raw material to deliverable < 30 min; loops completed per week; external MCP clients acting on real documents | |
There was a problem hiding this comment.
For the meeting loop highlighted in this row, the raw material is 45 minutes of live audio and the promised deliverable is complete when that meeting ends, so an end-to-end raw-material-to-deliverable target below 30 minutes cannot be applied consistently: measuring from capture start is impossible, while starting after capture conflicts with the already-complete result. Scope this target to report/deck loops or define it as processing/finalization lag so Phase 2 experiments have an unambiguous success criterion.
Useful? React with 👍 / 👎.
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: 6978d35790
ℹ️ About Codex in GitHub
Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you
- Open a pull request for review
- Mark a draft as ready
- Comment "@codex review".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".
| **Sequencing:** the phase opens with the cheapest loop (templates, already in | ||
| flight), then the keystone demo — *a 45-minute meeting ends with a complete, | ||
| local minutes document, decisions and action items with owners, captured | ||
| without note-taking* — then Report → Deck. The MCP server builds in parallel |
There was a problem hiding this comment.
This newly makes live 45-minute capture/STT the keystone immediately after templates, but docs/product/agent-native-priorities.md:44-52 still directs contributors to start with transcript import and add live audio/STT only after the proposal flow is stable. Contributors following that implementation plan will take the opposite critical path, so update it or explicitly mark it superseded as part of this roadmap reorder.
Useful? React with 👍 / 👎.
| | 1 | download → first agent-assisted document < 10 min; CI green ≥95% of days; ≥5 external PRs merged | | ||
| | Trusted Agent Actions | share of agent mutations accepted vs reverted (target ≥70% after the first iteration); time from opening a document to the first agent action | | ||
| | 2 | keystone demo ships (external agent edits a deck through the trust pipeline); external MCP clients acting on real documents; artifacts generated per agent session | | ||
| | 2 | keystone demo ships (45-min meeting → complete local minutes docx); report → deck pipeline; time from raw material to deliverable < 30 min; loops completed per week; external MCP clients acting on real documents | |
There was a problem hiding this comment.
For the meeting loop highlighted in this row, the raw material is 45 minutes of live audio and the promised deliverable is complete when that meeting ends, so an end-to-end raw-material-to-deliverable target below 30 minutes cannot be applied consistently: measuring from capture start is impossible, while starting after capture conflicts with the already-complete result. Scope this target to report/deck loops or define it as processing/finalization lag so Phase 2 experiments have an unambiguous success criterion.
Useful? React with 👍 / 👎.
…y notices (criptogus#52) SRC_GLOBS hardcodes five app source trees and enumerates packages/* dynamically, so apps/markdown was never scanned. The app ships: shell's electron-builder config copies ../markdown/out into every installer, and npm run notices is a prerequisite of dist:mac, dist:win and dist:linux. So the notices file that ships as extraResources omits the licenses of everything only markdown imports. Scanning it alone is not enough. markdown pins @tiptap 3.29.2 while docs takes ^3.28.0, so npm keeps markdown's copies under its own workspace, and starter-kit's own extensions land under a scoped package. pkgDir looked in neither: it iterated the entries of the root node_modules, which for a scoped package is the scope directory rather than the package, and never looked at workspace roots. Adds both lookups, unscoped owners first so packages that already resolved keep resolving to the same copy. Net effect on the generated file: 29 npm packages added, none removed, no version changed, and the "not installed, skipped" warning is now empty.
1 participant
What changes
Reorders Phase 2 around complete vertical loops (user goes from raw material → deliverable in one session) instead of horizontal agent infrastructure.
Main changes
Rationale
Each loop (meeting → minutes → action items → tasks; numbers → deck → approval → send; template → draft → deliverable) is a complete vertical the user feels daily — the retention story, and later the business story. The test for every item: does the user close the loop in one session?