SDK context integrity and merge truth by unbraind · Pull Request #652 · unbraind/pm-cli

unbraind · 2026-07-23T05:38:12Z

Summary

expose notes_count in token-bounded get/show projections while keeping deep/full note payloads backward compatible
promote alias-aware relationship edge validation into the public SDK and enforce self-edge policy before CLI create/update persistence
classify merge reconciliation by net outcome so append unions and reorderings preserve context without false data-loss warnings
document the new SDK and merge-safety contracts, generate the changelog through pm-changelog, and record rich tracker lineage

Why

Project management is context management. Agents need compact reads that disclose omitted context, mutation primitives that prevent invalid graph state, and merge diagnostics that stay loud only when context is genuinely lost.

Tracked work

pm-3esl28 — truthful omitted-note counts
pm-k9t17l — SDK relationship mutation integrity
pm-mmm9o5 — loss-aware merge reconciliation
pm-ds3b — rolling SDK reference

Validation

pnpm test:coverage — 6,101 tests; 100% statements, branches, functions, and lines
pnpm typecheck
pnpm lint — ESLint, zero-clone jscpd, static quality, 100% docstrings, package boundaries, and token budgets
docs skills and link validation
generated changelog check, secrets/version gates, packed npm smoke, and 88-command dogfood
isolated packed-consumer acceptance for CLI and public SDK behavior
real two-branch Git merge acceptance retaining both agents' notes with strict history validation

The consolidated release gate is otherwise green; its live Sentry check intentionally remains red on historical issue PM-CLI-2J from release 2026.7.22. This PR fixes that exact pre-persistence self-edge failure. The issue will remain open until the merged release is installed and verified with no recurrence.

Summary by cubic

Adds truthful SDK reads and safer writes: standard-depth get/show now reports omitted note counts, relationship mutations reject self-edges before persistence, and merge reconciliation only warns on real data loss.

New Features
- Standard-depth pm get/show --json returns item.notes_count when notes are omitted; deep/full payloads unchanged. (pm show --json (default 'standard' depth) silently omits notes and gives no notes_count — agents miss working context #645)
- SDK enforces relationship self-edge rules before persistence; CLI create/update reject self blocked_by and alias forms, honoring custom allow-self kinds. (Agent UX: pm update --dep accepts a self-referential ordering edge (id=<self>,kind=blocked_by) that pm graph audit forbids ('Self relationship is not allowed') #649)
- Merge reconciliation classifies discards as net-lossless vs net-lossy; unions/reorders keep attribution and avoid false loss warnings; docs updated. (Agent UX: pm merge reconcile over-reports discarded_authors/discarded_events on the driver union path (false data-loss alarm; complements #603) #646)
Migration
- Remove any automation that writes self blocked_by edges; these now error before write with a clear message.
- No changes needed for readers; notes_count is additive.

^{Written for commit f1b4c36. Summary will update on new commits.}

Expose bounded note counts, enforce relationship self-edge policy before persistence, and distinguish preserved merge unions from genuine reconciliation loss. Document the SDK contracts, add full acceptance coverage, and record the linked pm lineage and generated changelog.

gemini-code-assist · 2026-07-23T05:38:15Z

Caution

The consumer version of Gemini Code Assist on GitHub has been sunset. All code review activity has officially ceased.

sourcery-ai

Sorry @unbraind, you have reached your weekly rate limit of 500000 diff characters.

Please try again later or upgrade to continue using Sourcery

coderabbitai · 2026-07-23T05:38:20Z

Warning

Review limit reached

@unbraind, you've reached your PR review limit, so we couldn't start this review.

Next review available in: 17 minutes

Enable usage-based reviews in Billing to review now. Otherwise, wait until the next included review is available.
You're only billed for reviews past your plan's rate limits ($0.25/file).

How can I continue?

After more reviews become available, a review can be triggered using the @coderabbitai review command as a PR comment. Alternatively, push new commits to this PR.

To avoid repeated limits, reduce automatic review volume by pausing incremental auto-reviews earlier, using label-based review opt-in, excluding WIP or generated PR titles, or requesting reviews manually when the PR is ready. If your team needs uninterrupted high-volume reviews, an organization admin can enable usage-based reviews.

How do review limits work?

CodeRabbit enforces per-developer PR review limits for each organization. Most developers receive the normal plan review availability.

For paid Pro and Pro+ PR reviews, CodeRabbit uses adaptive limits for sustained high-volume activity. When a developer's recent PR review activity reaches the 95th percentile or higher among CodeRabbit users, additional reviews become available more gradually as earlier reviews age out of the rolling window.

Please refer docs for additional details.

Review details

⚙️ Run configuration

Configuration used: Repository UI (base), Organization UI (inherited)

Review profile: ASSERTIVE

Plan: Pro Plus

Run ID: 5bbbac6c-7a6e-4160-b4fd-a38a95dc3664

📥 Commits

Reviewing files that changed from the base of the PR and between 47f7f1f and f1b4c36.

📒 Files selected for processing (25)

.agents/pm/history/pm-3esl28.jsonl
.agents/pm/history/pm-7zs0.jsonl
.agents/pm/history/pm-ds3b.jsonl
.agents/pm/history/pm-h90s.jsonl
.agents/pm/history/pm-k9t17l.jsonl
.agents/pm/history/pm-mmm9o5.jsonl
.agents/pm/issues/pm-3esl28.toon
.agents/pm/issues/pm-h90s.toon
.agents/pm/issues/pm-k9t17l.toon
.agents/pm/issues/pm-mmm9o5.toon
.agents/pm/reminders/pm-7zs0.toon
.agents/pm/tasks/pm-ds3b.toon
CHANGELOG.md
docs/MERGE_SAFETY.md
docs/SDK.md
src/cli/commands/create.ts
src/cli/commands/get.ts
src/cli/commands/update.ts
src/sdk/dependency-flag-validation.ts
src/sdk/history-repair.ts
src/sdk/relationships.ts
tests/unit/commands/agent-ux-feedback-wave.spec.ts
tests/unit/commands/get-append-command.spec.ts
tests/unit/sdk/history-repair-reconciliation.spec.ts
tests/unit/sdk/relationships.spec.ts

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

_{Comment @coderabbitai help to get the list of available commands.}

unbraind · 2026-07-23T05:38:46Z

@greptileai review

@coderabbitai full review

Final review request for head 8f6ba73 after the tracker-link follow-up commit.

gemini-code-assist · 2026-07-23T05:38:46Z

Caution

The consumer version of Gemini Code Assist on GitHub has been sunset. All code review activity has officially ceased.

sourcery-ai

Sorry @unbraind, you have reached your weekly rate limit of 500000 diff characters.

Please try again later or upgrade to continue using Sourcery

coderabbitai · 2026-07-23T05:38:55Z

@unbraind I’ll perform a full review of PR #652 at head 8f6ba731, including the tracker-link follow-up.

✅ Action performed

Full review finished.

Your included review limit is currently reached under our Fair Usage Limits Policy. This review may still proceed through usage-based billing if eligible. Your next included review will be available in 27 minutes.

codspeed · 2026-07-23T05:40:55Z

Merging this PR will regress 3 benchmarks

⚡ 7 improved benchmarks
❌ 3 regressed benchmarks
✅ 1 untouched benchmark

Warning

Please fix the performance issues or acknowledge them on CodSpeed.

Performance Changes

	Benchmark	`BASE`	`HEAD`	Efficiency
❌	`createQueryFingerprint`	514.5 µs	515 µs	-0.08%
❌	`tokenizeAlphaNumeric`	367.2 µs	367.3 µs	-0.01%
❌	`score query over 2000 docs`	5.6 ms	5.6 ms	0%
⚡	`stableStringify (nested contract)`	161.3 µs	161.2 µs	+0.05%
⚡	`encodeQueryCursor`	101.2 µs	101.2 µs	+0.03%
⚡	`decodeQueryCursor`	116.1 µs	116.1 µs	+0.03%
⚡	`build index (2000 docs x 80 words)`	161.4 ms	161.4 ms	+0.03%
⚡	`tokenize + score query over 2000 docs`	7.5 ms	7.5 ms	+0.01%
⚡	`jaccardSimilarity`	153.3 µs	153.3 µs	+0.01%
⚡	`build index (200 docs x 40 words)`	8.4 ms	8.4 ms	+0.01%

Tip

Investigate this regression by commenting @codspeedbot fix this regression on this PR, or directly use the CodSpeed MCP with your agent.

_{Comparing feat/sdk-context-integrity-merge-truth (f1b4c36) with main (47f7f1f)}

codecov · 2026-07-23T05:41:39Z

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ All tests successful. No failed tests found.

📢 Thoughts on this report? Let us know!

unbraind · 2026-07-23T05:43:26Z

Acknowledged Gemini status report 5054869483: this reviewer surface is retired, so no code finding or action is present.

unbraind · 2026-07-23T05:43:27Z

Acknowledged CodeRabbit report 5054869925: the included review quota is temporarily exhausted; this is an external availability limit, not a code finding. Exact-head hosted gates remain green.

unbraind · 2026-07-23T05:43:28Z

Acknowledged Gemini status report 5054872526 on the final review request: the retired reviewer produced no code finding.

unbraind · 2026-07-23T05:43:29Z

Acknowledged CodeRabbit command response 5054873404: the final head was accepted for review but external fair-usage limits prevented findings. No code action is available from this response.

unbraind · 2026-07-23T05:43:30Z

CodSpeed report 5054884465 reviewed: the three flagged changes are 0 to 0.12 percent on untouched benchmark implementations, eight benchmarks improved, and the report warns that base/head used different runtime environments. This is non-actionable measurement noise; the benchmark workflow itself passed.

unbraind · 2026-07-23T05:43:31Z

Codecov report 5054888953 acknowledged: every modified coverable line is covered and all tests passed, consistent with the local and hosted 100/100/100/100 gates.

unbraind · 2026-07-23T05:43:33Z

Sourcery review PRR_kwDORSgRDc8AAAABG8gxKA acknowledged: weekly external diff quota exhausted; no code finding was produced.

unbraind · 2026-07-23T05:43:34Z

Sourcery exact-head review PRR_kwDORSgRDc8AAAABG8g7qQ acknowledged: weekly external diff quota exhausted; no code finding was produced at 8f6ba73.

unbraind · 2026-07-23T05:48:31Z

@greptileai review

@coderabbitai full review

Final governance-only follow-up at the new head records exact-hosted-review and scheduled-release evidence in pm. No product, test, or workflow code changed.

coderabbitai · 2026-07-23T05:48:41Z

@unbraind I’ll perform a full review of #652 at the latest head, including the governance-only tracker and scheduled-release evidence follow-up.

✅ Action performed

Full review finished.

Your included review limit is currently reached under our Fair Usage Limits Policy. This review may still proceed through usage-based billing if eligible. Your next included review will be available in 17 minutes.

unbraind · 2026-07-23T05:53:11Z

CodeRabbit final-head response 5054929513 acknowledged: f1b4c36 was accepted for full review, but the external fair-usage window still prevented findings. The governance-only push has 29 successful / 0 failing checks and no review threads.

unbraind · 2026-08-04T08:57:25Z

Tracked in PM: pm-3esl28, pm-ds3b, pm-k9t17l, pm-mmm9o5.

sourcery-ai Bot reviewed Jul 23, 2026

View reviewed changes

docs(pm): link SDK context integrity PR

8f6ba73

unbraind marked this pull request as ready for review July 23, 2026 05:38

sourcery-ai Bot reviewed Jul 23, 2026

View reviewed changes

unbraind mentioned this pull request Jul 23, 2026

Auto Release blocked: scheduled run failed #653

Closed

docs(pm): record review and release recovery

f1b4c36

unbraind merged commit 0c2928f into main Jul 23, 2026
32 checks passed

unbraind deleted the feat/sdk-context-integrity-merge-truth branch July 23, 2026 05:53

unbraind mentioned this pull request Jul 23, 2026

Record v2026.7.23 SDK context release proof #654

Merged

Conversation

unbraind commented Jul 23, 2026 • edited by cubic-dev-ai Bot Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Summary

Why

Tracked work

Validation

Summary by cubic

Uh oh!

gemini-code-assist Bot commented Jul 23, 2026

Uh oh!

sourcery-ai Bot left a comment

Choose a reason for hiding this comment

Uh oh!

coderabbitai Bot commented Jul 23, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Review limit reached

Uh oh!

unbraind commented Jul 23, 2026

Uh oh!

gemini-code-assist Bot commented Jul 23, 2026

Uh oh!

sourcery-ai Bot left a comment

Choose a reason for hiding this comment

Uh oh!

coderabbitai Bot commented Jul 23, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

codspeed Bot commented Jul 23, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Merging this PR will regress 3 benchmarks

Performance Changes

Uh oh!

codecov Bot commented Jul 23, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Codecov Report

Uh oh!

unbraind commented Jul 23, 2026

Uh oh!

unbraind commented Jul 23, 2026

Uh oh!

unbraind commented Jul 23, 2026

Uh oh!

unbraind commented Jul 23, 2026

Uh oh!

unbraind commented Jul 23, 2026

Uh oh!

unbraind commented Jul 23, 2026

Uh oh!

unbraind commented Jul 23, 2026

Uh oh!

unbraind commented Jul 23, 2026

Uh oh!

unbraind commented Jul 23, 2026

Uh oh!

coderabbitai Bot commented Jul 23, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

unbraind commented Jul 23, 2026

Uh oh!

Uh oh!

unbraind commented Aug 4, 2026

Uh oh!

Reviewers

Uh oh!

Assignees

Labels

Projects

Milestone

Development

Uh oh!

1 participant

unbraind commented Jul 23, 2026 •

edited by cubic-dev-ai Bot

Loading

coderabbitai Bot commented Jul 23, 2026 •

edited

Loading

coderabbitai Bot commented Jul 23, 2026 •

edited

Loading

codspeed Bot commented Jul 23, 2026 •

edited

Loading

codecov Bot commented Jul 23, 2026 •

edited

Loading

coderabbitai Bot commented Jul 23, 2026 •

edited

Loading