Back to Plate

tighten sync shadcn status output

docs/plans/2026-05-28-tighten-sync-shadcn-status-output.md

53.0.815.0 KB
Original Source

tighten sync shadcn status output

Objective: Tighten sync-shadcn status so routine status output reports only current sync state, partial syncs, deferred decisions, and next action. It must not list settled exclusions or preserved forks, and it must collapse Rhea/style/theme generated registry noise into the owning /create or theming product surface.

Goal plan: docs/plans/2026-05-28-tighten-sync-shadcn-status-output.md

Template: docs/plans/templates/task.md

Primary template: docs/plans/templates/task.md

Applied packs:

  • agent-native (docs/plans/templates/packs/agent-native.md)

Task source:

  • type: user correction
  • id / link: chat request
  • title: tighten sync-shadcn status
  • acceptance criteria: status no longer lists Settled exclusions or Preserved Plate forks; Rhea/style/theme generated registry rows are treated as create/theming product noise unless explicitly reviewed.

Completion threshold:

  • Complete only when .agents/rules/sync-shadcn.mdc documents the narrower status output and forbids listing settled exclusions or preserved forks.
  • Complete only when generated .agents/skills/sync-shadcn/SKILL.md is synced by pnpm install.
  • Complete only when focused audits prove the generated status output shape has no Settled exclusions or Preserved Plate forks sections, while retaining deferred-item guidance and source metadata.
  • Complete only when pnpm lint:fix, git diff --check, and node .agents/rules/autogoal/scripts/check-complete.mjs docs/plans/2026-05-28-tighten-sync-shadcn-status-output.md pass.

Verification surface:

  • pnpm install
  • source audits over .agents/rules/sync-shadcn.mdc, .agents/skills/sync-shadcn/SKILL.md, and .claude/skills/sync-shadcn/SKILL.md
  • focused sed read of the generated status output shape
  • pnpm lint:fix
  • git diff --check
  • autogoal completion checker

Constraints:

  • Edit .agents/rules/sync-shadcn.mdc as source of truth.
  • Do not manually edit generated skill mirrors.
  • Do not run build:registry.
  • Do not patch apps/www; this is a skill/command-contract change only.
  • Do not create PRs, commits, pushes, or tracker comments.

Boundaries:

  • Source of truth: .agents/rules/sync-shadcn.mdc
  • Allowed edit scope: .agents/rules/sync-shadcn.mdc, generated .agents/skills/sync-shadcn/SKILL.md, generated Claude mirror if Skiller updates it, and this goal plan.
  • Browser surface: N/A, no browser-visible app behavior changed.
  • Tracker sync: N/A, chat correction only.
  • Non-goals: run a new status command, implement a script runner, change sync state, or alter the full review output.

Output budget strategy:

  • Scope reads to the relevant status section and generated mirrors.
  • Use focused rg/sed; do not stream full run artifacts or shadcn diffs.

Blocked condition:

  • Block only if Skiller generation fails, generated mirrors cannot be synced from the source rule, or the status section cannot be made unambiguous without conflicting with review.

Task state:

  • task_type: agent command-contract correction
  • task_complexity: normal
  • current_phase: closeout
  • current_phase_status: complete
  • next_phase: final response
  • goal_status: active until completion checker and update_goal

Current verdict:

  • verdict: status contract corrected
  • confidence: high
  • next owner: sync-shadcn
  • reason: generated status output shape now contains only partial syncs, deferred decisions, and next action.

Completion rule:

  • Do not call update_goal(status: complete) while any checklist item remains unchecked.
  • Do not call update_goal(status: complete) until verification evidence is recorded and node .agents/rules/autogoal/scripts/check-complete.mjs docs/plans/2026-05-28-tighten-sync-shadcn-status-output.md passes.
  • Do not create hook state for this goal.

Start Gates:

GateAppliesEvidence
Skill analysis before editsyesLoaded sync-shadcn and autogoal; read current generated status section before editing.
Active goal checked or createdyesget_goal returned no active goal; create_goal created this correction objective.
Source of truth read before editsyes.agents/skills/sync-shadcn/SKILL.md identified source metadata; source rule edited.
Tracker comments and attachments readN/AChat correction only.
Video transcript evidence requiredN/ANo video/screenshot source.
docs/solutions checked for non-trivial existing-code workN/ASkill command text correction only.
TDD decision before behavior change or bug fixN/ANo executable behavior changed.
Branch decision for code-changing taskN/AUser did not ask for branch/PR work.
Release artifact decisionN/ANo package release artifact.
Browser tool decision for browser surfaceN/ANo browser surface changed.
PR expectation decisionN/AUser did not ask for PR.
Tracker sync expectation decisionN/ANo tracker target.
Output budget strategy recordedyesScoped reads/audits only.
Agent-native pack selectedyes.agents/** skill/rule surface changed.
Agent-facing action surface identifiedyessync-shadcn status output contract.
Source rule versus generated mirror boundary identifiedyesSource rule edited; Skiller regenerated mirrors via pnpm install.
agent-native-reviewer loaded or waiver recordedyesLoaded reviewer; no parity gap because the command itself is the agent action surface.

Work Checklist:

  • Objective includes outcome, completion threshold, verification surface, constraints, boundaries, and blocked condition.
  • Task source classified with source type, id/link, title, task type, acceptance criteria, caveats, likely files/routes/packages, browser surface, and root-cause layer.
  • Required video or screen-recording evidence is cached/read as normalized XML, or marked N/A with reason.
  • Nearby repo instructions and implementation patterns read before edits.
  • Implementation fixes the right ownership boundary, or the narrower choice is recorded with reason.
  • Release artifact requirement recorded: changeset, registry changelog, or N/A with reason.
  • Final handoff shape decided: concise correction summary, verification, and no PR/tracker lines.
  • Branch handling recorded for code-changing work: N/A, no branch/PR requested.
  • Local-env-rot retry policy recorded: N/A, no surprising repo-wide failure.
  • Workspace authority recorded: proof commands run in /Users/zbeyens/git/plate.
  • High-risk note recorded for command-contract change: failure mode is status output burying next actions under settled policy noise.
  • Review/autoreview target selected from actual diff state: agent-native reviewer loaded; autoreview N/A because this is docs/skill text only.
  • Agent-native review decision recorded for .agents/** skill changes.
  • Output budget discipline recorded and followed.
  • Agent-native pack: source-of-truth rule files are edited instead of generated skill mirrors.
  • Agent-native pack: the changed agent action is discoverable from the skill/rule text.
  • Agent-native pack: generated mirrors are synced when .agents/rules/** changed.
  • Agent-native pack: accepted agent-native review findings are fixed or explicitly rejected with reason.

Completion Gates:

GateAppliesRequired actionEvidence
Named verification thresholdyesProve status output is narrowed in source/generated skillFocused audits show generated status output has no Settled exclusions or Preserved Plate forks sections and keeps deferred guidance.
Bug reproduced before fixN/ARecord reasonUser corrected command-output shape, not a runtime bug.
Targeted behavior verificationyesAudit generated status sectionsed -n '148,168p' .agents/skills/sync-shadcn/SKILL.md shows only partial syncs, deferred decisions, and next action.
TypeScript or typed config changedN/ARecord reasonMarkdown rule/skill text only.
Package exports or file layout changedN/ARecord reasonNo package file layout or barrels changed.
Package manifests, lockfile, or install graph changedN/ARecord reasonpnpm install reported lockfile up to date; no package manifest edit for this task.
Agent rules or skills changedyesRun pnpm install and verify generated skill syncpnpm install passed and Skiller applied rules for Codex and Claude.
Workspace authority proofyesRun verification in owning repoAll commands ran from /Users/zbeyens/git/plate.
Browser surface changedN/ARecord waiverNo visible web UI changed.
Browser final proofN/ARecord waiverNo browser proof needed.
CI-controlled template output changedN/ARecord reasonNo registry/template output changed.
Package behavior or public API changedN/ARecord reasonSkill command contract only; no package changeset.
Registry-only component work changedN/ARecord reasonNo registry component work.
Docs or content changedN/ARecord reasonNo docs site content changed.
High-risk mini gateyesRecord failure mode, proof plan, and boundaryFailure mode: status lists settled policy instead of decisions; boundary: status contract in sync-shadcn; proof: generated output shape and forbidden-actions text.
Agent-native review for agent/tooling changesyesLoad reviewer and close findingsReviewer loaded; no actionable gap.
Local install corruption suspectedN/ARecord reasonNo install corruption signal.
Autoreview for non-trivial implementation changesN/ARecord reasonText-only skill contract; agent-native review is the relevant lane.
PR create or updateN/ARecord reasonUser did not ask for PR.
Task-style PR body verifiedN/ARecord reasonNo PR.
PR proof image hostingN/ARecord reasonNo PR/browser image.
Tracker sync-backN/ARecord reasonNo issue/Linear target.
Final handoff contractyesFill final handoff fieldsFilled below.
Final lintyesRun pnpm lint:fixPassed; no fixes applied.
Output budget disciplineyesVerify no unbounded outputAudits scoped to skill/rule paths and status section reads.
Goal plan completeyesRun completion checkerPassed.
Agent source / generated syncyesRun pnpm install and verify generated mirrorspnpm install passed; generated skill contains revised status section and source metadata.
Agent action discoverabilityyesSource-audit skill/rule pathGenerated skill status section is discoverable.
Agent-native reviewyesLoad reviewer and close findingsReviewer loaded; no findings.

Phase / pass table:

PhaseStatusEvidenceNext
Intake and source readcompleteSource/generated status sections and autogoal read.implementation
Implementationcomplete.agents/rules/sync-shadcn.mdc updated with narrowed status contract.verification
Verificationcompletepnpm install and focused generated-skill audits passed; final lint/checker next.closeout
PR / tracker syncN/ANo PR/tracker requested.final response
CloseoutcompleteFinal response after completion checker and goal close.final response

Findings:

  • User is right: Rhea/style/theme/generated style registry is upstream create/theming product infrastructure, not a useful standalone status item.
  • Routine status should not list settled exclusions or preserved forks. That is noise; review and the plan can hold that evidence.

Decisions and tradeoffs:

  • Keep status short and action-oriented.
  • Keep full policy/exclusion/fork evidence in review or plan artifacts.
  • Treat product/theme/style noise as part of the owning deferred product decision, not as a separate status line.

Implementation notes:

  • Removed settled-exclusion and preserved-fork sections from status output shape.
  • Removed partialSyncs[*].excluded from status summary inputs.
  • Added a rule to collapse upstream product/theme/style noise into its owning deferred decision.

Review fixes:

  • Agent-native reviewer loaded; no additional action required.

Error attempts:

Error / failed attemptCountNext different moveResolution
None0N/AN/A

Verification evidence:

  • pnpm install passed; Skiller applied rules for Codex and Claude.
  • Focused rg proved generated/source status text contains the new collapse rule, deferred decisions, status output shape, and source metadata.
  • sed -n '148,168p' .agents/skills/sync-shadcn/SKILL.md shows the generated status output shape without Settled exclusions or Preserved Plate forks.
  • rg -n 'Settled exclusions|Preserved Plate forks|partialSyncs[*].excluded' over source/Codex/Claude skill files returned no matches.
  • cmp showed Codex and Claude skill mirrors are identical; source differs from generated as expected because generated skill adds name and Skiller metadata.
  • pnpm lint:fix passed with no fixes applied.
  • node .agents/rules/autogoal/scripts/check-complete.mjs docs/plans/2026-05-28-tighten-sync-shadcn-status-output.md passed.

Final handoff contract:

  • PR line: N/A, no PR requested.
  • Issue / tracker line: N/A.
  • Confidence line: high.
  • Flow table:
    • Reproduced: N/A, requested command-output correction.
    • Verified: pnpm install, source audits, generated status-section audit, final lint, completion checker.
  • Browser check: N/A.
  • Outcome: sync-shadcn status no longer lists exclusions/forks and folds Rhea/style/theme generated registry noise into /create or theming product context.
  • Caveat: this updates the skill contract, not a script runner.
  • Design:
    • Chosen boundary: .agents/rules/sync-shadcn.mdc source rule.
    • Why not quick patch: editing generated SKILL.md directly would be overwritten and violate repo policy.
    • Why not broader change: review should retain full evidence; status should stay decision-focused.
  • Verified: final commands below.
  • PR body verified: N/A.

Task-style PR body contract:

  • N/A, no PR requested.

Final handoff / sync:

  • PR: N/A
  • Issue / tracker: N/A
  • Browser proof: N/A
  • Caveats: command contract only; no script runner added.

Timeline:

  • 2026-05-28T15:12:24.611Z Task goal plan created.
  • 2026-05-28 Tightened sync-shadcn status output contract.
  • 2026-05-28 Ran pnpm install to regenerate skill mirrors.

Reboot status:

QuestionAnswer
Where am I?Closeout
Where am I going?Final lint, completion checker, goal close
What is the goal?Tighten sync-shadcn status so it reports decisions, not settled policy noise.
What have I learned?Rhea/style/theme generated registry belongs under create/theming product context, not routine status.
What have I done?Updated source rule, regenerated skill mirrors, and audited generated output.

Open risks:

  • None for this task.