Core rich text operations AR perf

Objective: Optimize core-rich-text-operations-compare under Slate AR until target evidence is green, plateaued, or blocked by correctness/architecture proof.

Goal plan: docs/plans/2026-06-01-core-rich-text-operations-ar-perf.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)
• package-api (docs/plans/templates/packs/package-api.md)

Task source:

• type: user request
• id / link: local chat request, 2026-06-01
• title: Run Slate AR perf on core-rich-text-operations-compare
• acceptance criteria: initialize or resume the target-backed AR loop, run the rich-text compare benchmark, inspect the primary metric and correctness gate, optimize if a clear safe owner appears, and stop only when the metric is within target, plateaued, or blocked by a correctness/architecture reason.

Completion threshold:

• Done when core-rich-text-operations-compare has fresh target-backed AR evidence for rich_text_structural_ops_p95_ms, and one of these is true: metric is below the promotion target, at least two correctness-green packets produce less than 5% further gain, or the remaining owner is explicitly blocked by correctness/API architecture evidence. Promotion target from the registry is below 3x legacy on the structural composite.
• Task closure is legal only when the source-of-truth acceptance criteria are satisfied or explicitly narrowed, required verification evidence is recorded, code-review and release-artifact gates are closed when applicable, tracker/PR sync is complete or marked N/A with reason, and node .agents/skills/autogoal/scripts/check-complete.mjs docs/plans/2026-06-01-core-rich-text-operations-ar-perf.md passes.

Verification surface:

• Target setup: pnpm bench:targets:check, pnpm bench:targets:dry-run -- core-rich-text-operations-compare, and pnpm slate:ar:setup-target -- core-rich-text-operations-compare.
• Benchmark: RICH_TEXT_OPS_COMPARE_ITERATIONS=51 bun run bench:core:rich-text-operations:compare:local through the target or AR runner.
• Correctness: bun check in .tmp/slate-v2 when a packet is considered keep.
• Completion: autogoal completion check for this plan.

Constraints:

• Preserve existing user-facing behavior outside the task scope.
• Prefer the durable ownership boundary over caller-by-caller patches.
• Do not create PRs, comments, commits, or pushes unless the task/user/skill requires them.
• Do not add broad ceremony when the task is trivial or docs-only.

Boundaries:

• Source of truth: benchmarks/targets/slate-v2.json target core-rich-text-operations-compare plus .tmp/slate-v2/autoresearch.*.
• Allowed edit scope: .tmp/slate-v2 core/runtime files if profiling points to a safe owner, benchmark target/session artifacts, and this plan.
• Browser surface: N/A unless the target reveals a React/browser regression.
• Tracker sync: N/A, no issue/PR/tracker item requested.
• Non-goals: no pagination work in this loop; no review branch, commit, push, or PR unless explicitly requested.

Output budget strategy:

• Use target/AR commands and capped rg/sed reads. Keep large benchmark detail in JSON artifacts and report only metrics, deltas, and blockers.

Blocked condition:

• Block if AR refuses a safe segment because existing session/dirty-tree state cannot be switched, if bun check fails from unrelated existing work and the failing owner cannot be isolated, or if the next improvement requires a public API/architecture decision outside the current target.

Task state:

• task_type: performance autoresearch
• task_complexity: normal
• current_phase: closeout
• current_phase_status: complete
• next_phase: final response
• goal_status: active

Current verdict:

• verdict: green
• confidence: high
• next owner: no implementation owner
• reason: target-backed benchmark and two promotion-repeat packets are under the p95 ratio target with bun check passing.

Completion rule:

• Do not call update_goal(status: complete) while any required checklist item remains unchecked. If an item does not apply, check it and add N/A: <reason>.
• Do not call update_goal(status: complete) until every completion threshold above is satisfied, final handoff evidence is recorded, and node .agents/skills/autogoal/scripts/check-complete.mjs docs/plans/2026-06-01-core-rich-text-operations-ar-perf.md passes.
• Do not create hook state for this goal. This file plus the active goal are the durable state.

Start Gates:

Gate	Applies	Evidence
Skill analysis before edits	yes	Loaded autogoal, slate-ar-perf, and slate-ar.
Active goal checked or created	yes	get_goal returned none; created this goal.
Source of truth read before edits	yes	Read target core-rich-text-operations-compare from benchmarks/targets/slate-v2.json.
Tracker comments and attachments read	no	N/A: no tracker item or attachment.
Video transcript evidence required	no	N/A: no video evidence in this task.
docs/solutions checked for non-trivial existing-code work	no	N/A: no code owner was changed; target was already green.
TDD decision before behavior change or bug fix	no	N/A: no behavior patch was needed.
Branch decision for code-changing task	no	N/A: no branch/commit/PR requested.
Release artifact decision	yes	No release artifact: no runtime/package source change was kept.
Browser tool decision for browser surface	no	N/A unless a browser/React regression appears.
PR expectation decision	yes	No PR requested.
Tracker sync expectation decision	yes	No tracker sync requested.
Output budget strategy recorded	yes	Recorded above.
Agent-native pack selected	yes	Target/AR package scripts are agent-facing workflow surfaces.
Agent-facing action surface identified	yes	bench:targets:*, slate:ar:*, and .tmp/slate-v2 benchmark scripts.
Source rule versus generated mirror boundary identified	yes	Target registry/session files are source for this loop; no generated skill mirror touched.
agent-native-reviewer loaded or waiver recorded	no	N/A: no agent rule/skill/tool source changed; only existing AR session files were pointed at the selected target.
Package/API pack selected	yes	Possible package runtime perf changes in .tmp/slate-v2/packages/**.
Public surface or package boundary identified	yes	Potential package runtime only; no public API planned.
Release artifact path selected	yes	N/A: no published user-visible delta.
changeset skill loaded when .changeset is required	no	N/A: no changeset required.
Barrel/export impact decision recorded	no	N/A: no exports or file layout changed.

Work Checklist:

• Short objective plus outcome, completion threshold, verification surface, constraints, boundaries, and blocked condition are concrete.
• 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 <video-transcripts> 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: bug/feature/testing/batch/review/tracker requirements, PR body sync, and issue/Linear sync when applicable.
• Branch handling recorded for code-changing work: dedicated branch used, new branch needed, or N/A with reason.
• Local-env-rot retry policy recorded for any surprising repo-wide failure: reinstall/rerun evidence or N/A with reason.
• Workspace authority recorded: every proof command names the cwd/tool that owns the changed behavior.
• High-risk note recorded for public API, runtime, package-boundary, browser behavior, agent-action, or command-contract changes, or marked N/A with reason.
• Review/autoreview target selected from actual diff state for non-trivial implementation work, or marked N/A with reason.
• Agent-native review decision recorded for .agents/**, .claude/**, .codex/**, skills, hooks, commands, prompts, or user-action tooling.
• Output budget discipline recorded and followed: broad searches are scoped, capped, counted, or artifacted instead of streamed into goal context.
• 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, or N/A reason is recorded.
• Agent-native pack: accepted agent-native review findings are fixed or explicitly rejected with reason.
• Package/API pack: public API, package boundary, export, and release-artifact impact are recorded.
• Package/API pack: release artifact matrix is applied: .changeset, registry changelog, or explicit no-artifact reason.
• Package/API pack: .changeset work loads changeset and follows its package/version/prose rules.
• Package/API pack: registry-only work updates docs/components/changelog.mdx instead of adding a package changeset.
• Package/API pack: no-artifact decisions state why the diff has no published package user-visible delta from main.
• Package/API pack: compatibility, migration, or hard-cut decision is explicit when public shape changes.
• Package/API pack: package-owned typecheck/build/test proof is recorded or marked N/A with reason.
• Package/API pack: generated barrels or release notes are updated when required.

Completion Gates:

Gate	Applies	Required action	Evidence
Named verification threshold	yes	Run the target setup, benchmark, promotion repeat, checks, and state proof named in this plan.	Target setup OK; benchmark lint emitted primary metric; packet 4 was 1.03ms; promotion repeats were 1.6ms and 0.94ms; all packet checks passed.
Bug reproduced before fix	no	Record N/A with reason.	N/A: no product bug was fixed.
Targeted behavior verification	yes	Run target benchmark plus correctness gate.	pnpm slate:ar:next packets ran bash ./autoresearch.sh and bash ./autoresearch.checks.sh; checks passed each time.
TypeScript or typed config changed	no	Record N/A with reason.	N/A: no TS or typed config source changed.
Package exports or file layout changed	no	Record N/A with reason.	N/A: no exports or file layout changed.
Package manifests, lockfile, or install graph changed	no	Record N/A with reason.	N/A: no manifest, lockfile, or install graph changed.
Agent rules or skills changed	no	Record N/A with reason.	N/A: no agent source changed.
Workspace authority proof	yes	Run proof in the owning workspace.	plate-2 ran target commands; .tmp/slate-v2 ran benchmark/check wrappers.
Browser surface changed	no	Record N/A with reason.	N/A: no browser behavior changed.
Browser final proof	no	Record N/A with reason.	N/A: no browser proof required.
CI-controlled template output changed	no	Record N/A with reason.	N/A: no templates/** touched.
Package behavior or public API changed	no	Record N/A with reason.	N/A: no runtime patch was kept.
Registry-only component work changed	no	Record N/A with reason.	N/A: no registry component work.
Docs or content changed	yes	Verify source-backed plan claims.	This plan records exact command evidence; no docs claim depends on runtime rendering.
High-risk mini gate	yes	Record failure mode and proof plan.	Risk was optimizing wrong target due stale AR wrapper; benchmark-lint caught it, wrapper files were corrected, new segment was created, and promotion repeats passed.
Agent-native review for agent/tooling changes	no	Record N/A with reason.	N/A: no .agents/**, .claude/**, .codex/**, skill, hook, command, or prompt source changed.
Local install corruption suspected	no	Record N/A with reason.	N/A: no install corruption signal appeared.
Autoreview for non-trivial implementation changes	no	Record N/A with reason.	N/A: no product/runtime code patch was kept; session and plan changes are mechanically verified.
PR create or update	no	Record N/A with reason.	N/A: no PR requested.
Task-style PR body verified	no	Record N/A with reason.	N/A: no PR requested.
PR proof image hosting	no	Record N/A with reason.	N/A: no PR/browser proof image.
Tracker sync-back	no	Record N/A with reason.	N/A: no tracker item.
Final handoff contract	yes	Fill final handoff fields.	Final handoff fields below record outcome, caveat, design, and verification.
Final lint	no	Record N/A with reason.	N/A: no product code changed; benchmark/check wrappers executed successfully.
Output budget discipline	yes	Verify output stayed bounded.	Long benchmark output was summarized by AR tails and metrics; tool output was capped.
Goal plan complete	yes	Run node .agents/skills/autogoal/scripts/check-complete.mjs docs/plans/2026-06-01-core-rich-text-operations-ar-perf.md.	Run after this plan is filled.
Agent source / generated sync	no	Record N/A with reason.	N/A: no .agents/rules/** changed.
Agent action discoverability	yes	Source-audit action surface.	bench:targets:*, slate:ar:*, .tmp/slate-v2/autoresearch.sh, and .tmp/slate-v2/autoresearch.checks.sh identify the target loop.
Agent-native review	no	Record N/A with reason.	N/A: no agent source changed.
Public API / package boundary proof	yes	Source-audit boundary.	No public package API/export changed; no code patch kept.
Release artifact classification	yes	Record classification.	Internal AR/session/plan evidence only.
Published package changeset	no	Record N/A with reason.	N/A: no published package delta.
Registry changelog	no	Record N/A with reason.	N/A: no registry component work.
No release artifact	yes	Record exact reason.	Internal-only measurement/session update; no user-visible package behavior/API/types/config/runtime delta.
Package typecheck/build/test	yes	Run owning package check or N/A.	bun check ran as the AR correctness command for packets 4, 5, and 6 and passed.
Barrel/export generation	no	Record N/A with reason.	N/A: no exports changed.

Phase / pass table:

Phase	Status	Evidence	Next
Intake and source read	complete	Target contract and AR skills read.	implementation
Implementation	complete	Corrected stale AR session wrappers and started target/promotion segments.	verification
Verification	complete	Benchmark lint, packet 4, promotion repeats 5/6, and slate:ar:state passed.	closeout
PR / tracker sync	N/A	No PR or tracker sync requested.	final response
Closeout	complete	Plan filled; autogoal check runs last.	final response

Findings:

• pnpm slate:ar:init-target updated config but kept stale pagination wrappers; benchmark-lint caught this by emitting pagination metrics instead of rich_text_structural_ops_p95_ms.
• After correcting wrappers and starting a new segment, the target emitted the intended metric and passed the correctness gate.
• Promotion repeats were green: 1.6ms / p95 ratio 2.36, then 0.94ms / p95 ratio 1.85; both passed bun check.

Decisions and tradeoffs:

• Logged benchmark packets as measure, not keep, because there was no product patch to commit. keep would have been fake here.
• Stopped after the promotion-repeat gate because the target is under the p95 ratio threshold; further tuning would chase noise.

Implementation notes:

• Updated .tmp/slate-v2/autoresearch.sh to run RICH_TEXT_OPS_COMPARE_ITERATIONS=51 bun run bench:core:rich-text-operations:compare:local.
• Updated .tmp/slate-v2/autoresearch.checks.sh to run bun check.
• Updated .tmp/slate-v2/autoresearch.md so the session text matches the core rich-text target.

Review fixes:

• None: no product code patch.

Error attempts:

Error / failed attempt	Count	Next different move	Resolution
Initial benchmark-lint emitted pagination metrics	1	Patch stale AR wrapper files and start a new rich-text segment	Resolved; benchmark-lint then emitted rich_text_structural_ops_p95_ms=0.98

Verification evidence:

• pnpm bench:targets:check: benchmark-targets ok: 26 targets.
• pnpm bench:targets:dry-run -- core-rich-text-operations-compare: setup OK for rich_text_structural_ops_p95_ms.
• pnpm slate:ar:setup-target -- core-rich-text-operations-compare: detected current metric was still pagination and recommended setup.
• pnpm slate:ar:init-target -- core-rich-text-operations-compare: initialized rich-text config.
• First pnpm slate:ar:benchmark-lint: failed because stale wrapper emitted pagination metrics; fixed wrapper/session files.
• node ... autoresearch.mjs new-segment --cwd .tmp/slate-v2 --reason "Switch active target from pagination to core-rich-text-operations-compare" --yes: segment 2 created.
• Second pnpm slate:ar:benchmark-lint: OK, rich_text_structural_ops_p95_ms=0.98, worst p95 ratio 2.46, worst mean ratio 2.27.
• pnpm slate:ar:next packet 4: 1.03ms, worst p95 ratio 2.35, worst mean ratio 3.17, bun check passed.
• Logged packet 4 as accepted measure; no commit created.
• node ... promote-gate --cwd .tmp/slate-v2 --reason "Core rich-text baseline is below the <3x legacy p95 promotion target" --gate-name "rich-text promotion repeat" --query-count 2 --yes: promotion segment 3 created.
• Promotion packet 5: 1.6ms, worst p95 ratio 2.36, worst mean ratio 3.75, bun check passed; logged accepted measure.
• Promotion packet 6: 0.94ms, worst p95 ratio 1.85, worst mean ratio 1.92, bun check passed; logged accepted measure.
• pnpm slate:ar:state: active target core-rich-text-operations-compare; segment 3 has 2 measured, 0 checks failed; researchIntegrity OK; next hint says close target as green.

Final handoff contract:

• PR line: N/A, no PR requested.
• Issue / tracker line: N/A, no tracker item.
• Confidence line: high for target green; no runtime patch was needed.
• Flow table:
  • Reproduced: N/A, no product bug.
  • Verified: benchmark lint, packet 4, two promotion repeats, bun check in each packet, and slate:ar:state.
• Browser check: N/A, no browser surface changed.
• Outcome: core-rich-text-operations-compare is green under the target's p95 ratio threshold.
• Caveat: worst mean ratio was noisy in packets 4/5, but promotion packet 6 brought p95 and mean ratios both under 2x; the primary target is p95.
• Design:
  • Chosen boundary: AR target/session proof, not runtime code.
  • Why not quick patch: metric already passes; code edits would be noise.
  • Why not broader change: this target is scoped to rich-text structural ops, not pagination or full all-lane benchmarking.
• Verified: see verification evidence.
• PR body verified: N/A, no PR.

Task-style PR body contract:

• Preserve any existing <!-- auto-release:start --> block. If a changeset is part of the diff and repo policy expects auto release, include that block.
• Use the accepted kitcn PR #270 visual format. The body starts with an emoji issue/tracker/fix line, for example 🐛 Fixes #123 or 🐛 Fixes ➖ N/A, then an emoji confidence line like 🟢 95-100% confidence.
• Use this exact table header: | Phase | 🧪 Tests | 🌐 Browser |.
• Use Reproduced and Verified rows. Mark passing proof with 🟢, repro or failing proof with 🔴, and non-applicable cells with ➖ N/A.
• Use bold emoji section headings: **✅ Outcome**, **⚠️ Caveat**, **🏗️ Design**, and **🧪 Verified**.
• Never include a line that links to the current PR itself. The current PR URL belongs in the final response, not in its own description.
• Do not replace this with a generic Summary / Verification PR body, an adaptive prose body from a git helper skill, plain ## Outcome sections, or an unrelated generated badge footer unless the caller or repo template explicitly asks for it.
• Proof is gh pr view --json body output or a concise source-backed summary of that output.

Final handoff / sync:

• PR: N/A, no PR requested.
• Issue / tracker: N/A, no tracker item.
• Browser proof: N/A, no browser surface changed.
• Caveats: no code patch; AR session files changed to point at this target.

Timeline:

• 2026-06-01T16:07:28.588Z Task goal plan created.
• 2026-06-01T16:11Z Rich-text AR session segment created after wrapper fix.
• 2026-06-01T16:12Z Packet 4 measured 1.03ms with bun check green.
• 2026-06-01T16:15Z Promotion gate segment created with 2 repeats.
• 2026-06-01T16:16Z Packet 5 measured 1.6ms with bun check green.
• 2026-06-01T16:17Z Packet 6 measured 0.94ms with bun check green.

Reboot status:

Question	Answer
Where am I?	Closeout
Where am I going?	Final response after completion check
What is the goal?	Optimize or prove green core-rich-text-operations-compare under Slate AR
What have I learned?	It is already green under p95 target; stale AR wrappers were the only setup bug
What have I done?	Fixed target wrappers, ran baseline and promotion repeats, logged accepted measurements

Open risks:

• None for this target. Further optimization would be benchmark-noise chasing unless a future all-lane sweep exposes a regression.