Skip to content

spec(FIX-1903): end-state figure shows FSD holding the MCP connection with each person's key - #3448

Merged
jhoffner merged 2 commits into
mainfrom
claude/project-thread-exyje0
Oct 15, 2026
Merged

jhoffner merged 2 commits into
mainfrom
claude/project-thread-exyje0

Conversation

@jhoffner

@jhoffner jhoffner commented Oct 14, 2026 •

Copy link
Copy Markdown
Collaborator

Requested by Jake · project thread

Before: the FIX-1903 end-state figure (specs/epics/FIX-1903/figures/end-state.svg) said a project's MCP endpoint and credential live in host config, and its "not built" fence listed "per-person Linear sign-in". Both contradict amendment #3179, which the #3179 Codex review flagged (P2).

After: the figure matches D4 as amended. The row names a host alias only (no URL, no key), FSD connects for the run with the key each person sets on the Settings page, and the endpoint stays host config. The fence now says what is actually out of scope: Linear sign-in by OAuth.

How: one figure amendment from main. The MCP box gains one line ("FSD connects with your key (Settings)") and grows 6px to fit. The aria-label and the figure's alt text in SPEC.md say the same thing. Rendered in light and dark themes and checked; no other text changed. No changeset (spec only).

🤖 Generated with Claude Code

https://claude.ai/code/session_01PrPKdauAZJcDc2GUGxNKWS


Generated by Claude Code

RetriggerView in GreptileConfidence Score: 5/5

This spec-only correction appears safe to merge.

Summary

Updates the end-state figure to match the amended MCP design.

  • The end-state figure shows FSD connecting with each person’s Settings key.

Reviews (1) · Last reviewed commit: "spec(FIX-1903): end-state figure shows F..." · Reviewed by Greptile

… with each person's own key

The figure still put the run's credential in host config. Amendment #3179
settled that FSD holds the MCP connection and each person sets their own
key on the Settings page; the row names a host alias only.

Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01PrPKdauAZJcDc2GUGxNKWS
@linear-code

linear-code Bot commented Oct 14, 2026

Copy link
Copy Markdown

FIX-1903

@vercel

vercel Bot commented Oct 14, 2026 •

Copy link
Copy Markdown

The latest updates on your projects. Learn more about Vercel for GitHub.

3 Skipped Deployments
Project Deployment Actions Updated
flow-state-dev-docs Ignored Ignored Oct 15, 2026 12:00am UTC
flow-state-dev-kitchen-sink Ignored Ignored Oct 15, 2026 12:00am UTC
flow-state-dev-ui Ignored Ignored Preview Oct 15, 2026 12:00am UTC

Request Review

@jhoffner jhoffner self-assigned this Oct 14, 2026

Copy link
Copy Markdown
Collaborator Author

Second look / architecture pass: no changes requested.

  • Approach: a figure-and-alt-text amendment is the right size. The figure only had to agree with D4 as amended (spec(FIX-1903): epic amendment, FSD holds MCP keys per user #3179), and a leaner fix doesn't exist. I found no refactor opportunity tied to this change.
  • Consistency: the new fence wording ("Linear sign-in by OAuth") matches SPEC.md:31 ("signing in to Linear by OAuth, each person pastes their own key instead"). The box text matches D4's "Locks in" (DECISIONS.md:81) and ER-6 (no URL or key on the row). The aria-label, the SPEC.md alt text and the visible SVG text all say the same thing. A grep of the other FIX-1903 figures found no remaining "credential in host config" claim.
  • Layout: both held boxes grow to 92px, so they stay matched. The new baseline at y=218 fits inside the box (ends at y=228) and the fence at y=256 is untouched.
  • Nit, optional: the box says "your key" and the alt text says "the key each person sets". Both are accurate, so leave it as is.

Generated by Claude Code

@cursor cursor Bot left a comment

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Summary

This PR does the right thing at the right altitude: it corrects the end-state figure and alt text after the FIX-1905 review superseded “credential in host config” with endpoint in host config + per-person key on Settings + FSD holds the connection (ER-6 / EVOLUTION row on #3174). The OAuth fence rename (no Linear sign-in by OAuth) also lines up with the epic “Bigger” row and FIX-1905’s d4-whose-key flip condition better than “per-person Linear sign-in,” which read like in-scope sign-in.

I’m approving because the substance matches the locked rules; what’s left is figure/copy hygiene, not direction.

Simpler approach worth considering (not blocking)

The content fix is necessary; the shape of the fix may be heavier than the repo’s own layering pattern:

  1. Epic end-state = panorama; FIX-1905 what-changes.svg + ER-6 = credential mechanics. Other epics keep held boxes at ~3 sublines. Here the MCP box grew to five lines (+ taller rects) while d4-read-at-run-start.svg and FIX-1905 already say “FSD connects; harness gets tools” and the two-input diagram (host config vs Settings). A cleaner change might replace the wrong “credential in host config” line with one merged line (e.g. alias on row · endpoint in host config · FSD connects with your Settings key) instead of appending a fifth line—same truth, less layout churn.

  2. Two long prose copies. Markdown ![…] alt and SVG aria-label are both full essays and already diverge (NOT BUILT list, Layer 1, migration wording). Ongoing cost every time the figure moves. Simpler maintenance: short alt (“End-state: in the box, unchanged, replaced, not built”) and let aria-label (or <title>/<desc>) be canonical—or keep one list in sync, not two hand-edited narratives.

  3. Small clarity nits (if you touch the SVG again): “no key on it” → “on the row” (matches ER-6 / aria-label); unify “connected by FSD” (alt) vs “FSD connects” (visible) with D4’s active voice; optional alignment with SPEC line 31 parenthetical for OAuth (fence stays short; nuance in Bigger row).

What I’d skip

  • Duplicating FIX-1905’s topology inside the epic end-state—that would be the approach that went too far.
  • Third-party packages: N/A (spec SVG only).

Performance: N/A.

Open in Web View Automation 

Sent by Cursor Automation: Review For Best Approach

Comment thread specs/epics/FIX-1903/SPEC.md Resolved
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01PrPKdauAZJcDc2GUGxNKWS
@jhoffner

Copy link
Copy Markdown
Collaborator Author

@codex review

@chatgpt-codex-connector

chatgpt-codex-connector Bot commented Oct 15, 2026 •

Copy link
Copy Markdown

Codex Review Summary

This comment shows the latest Codex review activity on this pull request.

Review Status Commit Review trigger
📝 Code Review ✅ Completed 2026-10-15T01:30:58.217376Z ced9792 Manual request
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" or "@codex security review".

Codex reacts with 👀 while any review is running, comments if it has suggestions, and reacts with 👍 once all reviews finish with no findings.

@chatgpt-codex-connector

Copy link
Copy Markdown

Codex Review: Didn't find any major issues. More of your lovely PRs please.

Reviewed commit: ced9792695

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".

@jhoffner
jhoffner merged commit 334e79b into main Oct 15, 2026
18 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants