Commit Graph
2 Commits
Author SHA1 Message Date
a5e01634f6 feat(ui): ask self-hosted users to star Paperclip on GitHub after 5 done tasks (#20453)
## Thinking Path

> - Paperclip is the open source app people use to manage AI agents for
work
> - Many teams run Paperclip self-hosted, so the project grows when
those users tell others about it
> - A GitHub star is the cheapest way for a user to do that, but the app
never asks
> - Asking too early is noise. A user who has finished a few tasks knows
whether Paperclip helps them
> - This pull request adds a one-time, dismissable modal that asks
self-hosted users for a star after their company has 5 done tasks
> - The benefit is more visibility for the project, and no repeated
prompt for the user

## Linked Issues or Issue Description

**Subsystem affected**

Board UI (dialogs shown in the main app shell).

**Problem or motivation**

Self-hosted users have no prompt to star the repository. Users who like
Paperclip often do not think of it. A prompt at the right moment helps
other teams find the project.

**Proposed solution**

Show a small modal once the selected company has 5 tasks with status
"done". The modal shows the Paperclip mark, the headline "You completed
5 tasks!", a short body, a "Star on GitHub" button and a "Not now"
button. It shows only once per browser. Closing it in any way counts as
seen.

**Alternatives considered**

A permanent banner or sidebar link. Both stay on screen after the user
decides, so they cost attention every day. A one-time modal costs
attention once.

**Roadmap alignment**

This is a small UI addition. It does not overlap with planned core work
in ROADMAP.md.

No related public issues or PRs found (searched for "star", "github
star").

## What Changed

- Added `GitHubStarPrompt` (`ui/src/components/GitHubStarPrompt.tsx`).
It reads the health endpoint and shows only on self-hosted instances
(never on Cloud), never during onboarding, and only after 5 done tasks
in the selected company. It waits for any other open dialog to close. It
stores a "seen" flag in `localStorage` when it opens, so a reload does
not show it again.
- Mounted the prompt in `BoardDialogs`.
- Exported the modal-detection helpers (`MODAL_SELECTOR`,
`useModalOpen`) from `AnnouncementWell` so both components use the same
rule.
- Added `PaperclipMark`, the mark alone from the existing lockup
geometry. The lockup and the mark now share one path constant.
- Added the `issues.starPromptDone` query key.
- Added tests for when the prompt opens and when it stays hidden.

## Verification

- `pnpm --filter @paperclipai/ui exec vitest run
src/components/GitHubStarPrompt.test.tsx
src/components/AnnouncementWell.test.tsx`
- `pnpm check:token-gates` (all gates clean)
- `pnpm typecheck`
- Manual: on a self-hosted dev instance, mark 5 tasks in one company as
done. The modal opens. "Star on GitHub" opens
github.com/paperclipai/paperclip in a new tab. Reload: the modal does
not open again. To see it again, delete the
`paperclip.github-star-prompt.seen` key from `localStorage`.

## Risks

- Low risk. The change is UI only and additive.
- The prompt runs one small issues query (limit 5) while it is still
eligible. After it is seen, the query is disabled.
- If `localStorage` is not available, the prompt treats itself as
already seen and does not show.

> For core feature work, check [`ROADMAP.md`](ROADMAP.md) first and
discuss it in `#dev` before opening the PR. Feature PRs that overlap
with planned core work may need to be redirected — check the roadmap
first. See `CONTRIBUTING.md`.

## Model Used

- Claude Opus 5.5 (Anthropic), model id `claude-opus-5-5`, run through
Claude Code with tool use (shell, file edits). Standard context window;
extended thinking at low effort.

## Checklist

- [x] I have included a thinking path that traces from project context
to this change
- [x] I have specified the model used (with version and capability
details)
- [x] I have checked ROADMAP.md and confirmed this PR does not duplicate
planned core work
- [x] I have searched GitHub for duplicate or related PRs and linked
them above
- [x] I have either (a) linked existing issues with `Fixes: #` / `Closes
#` / `Refs #` OR (b) described the issue in-PR following the relevant
issue template
- [x] I have not referenced internal/instance-local Paperclip issues or
links (only public GitHub `#NNN` / `github.com/paperclipai/paperclip`
URLs)
- [x] My branch name describes the change (e.g. `docs/...`, `fix/...`)
and contains no internal Paperclip ticket id or instance-derived details
- [x] I have run tests locally and they pass
- [x] I have added or updated tests where applicable
- [x] I have updated relevant documentation to reflect my changes
- [x] I have considered and documented any risks above
- [ ] All Paperclip CI gates are green
- [ ] Greptile is 5/5 with no open P2s, recommendations, or follow-ups
- [x] I will address all Greptile and reviewer comments before
requesting merge

🤖 Generated with [Claude Code](https://claude.com/claude-code)

---------

Co-authored-by: squadbot <squadbot@squadbots-Mac-mini.local>
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-09 21:11:29 -08:00
TonioandClaude Opus 5 88a0f885e8 Brand lockup, no idle env-check card, no Mission row (#12074)
## Thinking Path

> - Paperclip is the open source app people use to manage AI agents for
work
> - New customers meet the product through an onboarding arc that ends
in the tenant wizard
> - A staging walk of that arc found three rough edges: an ad-hoc icon
standing in for the brand, an environment-check card that narrates a
probe the flow already runs on its own, and a review checklist that
still lists a Mission the arc stopped asking for
> - Each one makes the product look less finished than it is at the
exact moment a customer decides what it is
> - This pull request renders the brand lockup, hides the idle
environment-check card while keeping the probe and its failure surface,
and drops the Mission row
> - The benefit is a first-session arc that reads as one product, with
no controls for questions nobody was asked

## Linked Issues or Issue Description

No public issue exists. The changes come from walking the sign-up arc on
a staging fleet.

**What happened:**
The model step shows an "Adapter environment check" card with a "Test
now" button even though pressing Connect runs the same probe and blocks
a failing hire. The review step lists "Mission" in its checklist
although onboarding no longer asks for one. The auth page renders a
sparkles icon beside the word "Paperclip" instead of the brand lockup.

**Expected behavior:**
The model step shows the check only when a probe has found something to
fix. The review checklist lists only what onboarding set up. The brand
renders as the lockup asset used across surfaces.

**Actual behavior:**
An idle card narrates a probe that runs regardless. A permanent
unchecked row marks a question nobody was asked. The brand is a generic
icon plus text.

**Steps to reproduce:**
1. Sign up on a staging fleet and enter the tenant wizard.
2. On "Create your first agent", choose a role and press Next: the model
step shows the "Adapter environment check" card before anything has been
probed.
3. Continue to Review: the checklist lists "Mission" as a permanently
unchecked row.

**Additional context:**
The Mission row outlived the removal of the mission step (#11935). The
environment probe itself still runs on Connect and blocks a failing
hire; only its idle card is at issue. The brand lockup lands across all
three surfaces in the same round — paperclip-cloud#270 and
paperclip-id#58 carry the other halves.

## What Changed

- `PaperclipLockup` renders the brand asset (mark + wordmark, one
geometry, `fill="currentColor"`); the auth page uses it in place of the
sparkles icon.
- The adapter environment check's idle card (explainer + "Test now") no
longer renders. The probe still runs on Connect and still blocks a
failing hire.
- The check's failure content still renders when a probe has found
something — the blocking error points the customer at "the reported
checks", so they stay visible.
- Connect retries a cached failed probe instead of reusing it. With
"Test now" gone, Connect is the only retry, and a stale fail would lock
out a machine the customer has since fixed.
- The review checklist drops its "Mission" row.

## Verification

Run the tenant suite:

```
cd ui && npx vitest run
```

- 4365 tests pass across 471 files; `npx tsc --noEmit` is clean.
- New test drives the wizard to the model step and asserts the
environment-check card is absent, anchored on "Connect a model" so an
unrendered step cannot pass as an absence.
- The review assertions anchor on the remaining rows ("Organization
name", "Agent created", "Model connected").

## Risks

- **Behavioral change:** a cached *failed* probe is re-run on Connect
instead of reused. Pass and warn results are still cached. This only
affects the retry path that "Test now" used to serve.
- **Hidden, not removed:** the environment check machinery is intact;
only the idle card is gone. A failing probe still blocks the hire and
still shows its checks.
- **Brand:** the wordmark now ships inside an SVG; its accessible name
carries the text. Screen readers announce "Paperclip" as before.

## Model Used

Claude Fable 5 (`claude-fable-5`) via Claude Code, with tool use and
code execution; earlier rounds on this branch's predecessor used Claude
Opus 5 (`claude-opus-5`).

## Checklist

- [x] I have included a thinking path that traces from project context
to this change
- [x] I have specified the model used (with version and capability
details)
- [x] I have checked ROADMAP.md and confirmed this PR does not duplicate
planned core work
- [x] I have searched GitHub for duplicate or related PRs and linked
them above
- [x] I have either (a) linked existing issues with `Fixes: #` / `Closes
#` / `Refs #` OR (b) described the issue in-PR following the relevant
issue template
- [x] I have not referenced internal/instance-local Paperclip issues or
links
- [x] My branch name describes the change and contains no internal
Paperclip ticket id
- [x] I have run tests locally and they pass
- [x] I have added or updated tests where applicable
- [ ] I have updated relevant documentation to reflect my changes
- [x] I have considered and documented any risks above
- [ ] All Paperclip CI gates are green
- [ ] Greptile is 5/5 with no open P2s, recommendations, or follow-ups
- [x] I will address all Greptile and reviewer comments before
requesting merge

---------

Co-authored-by: Claude Opus 5 <noreply@anthropic.com>
2026-08-24 01:35:15 -07:00