Release ultimate-agent-stack 0.9.1

[samtay32]

Summary

• release the merged concise install/upgrade improvements as ultimate-agent-stack@0.9.1
• synchronize npm, plugin, CLI fallback, lockfile, and version-bound tests
• move the completed changes from Unreleased into the 0.9.1 changelog
• regenerate deterministic fixture baselines for the exact installed bytes

Why

npm already publishes 0.9.0 and immutable registry versions cannot be overwritten. The improvements merged in PR #36 therefore require a patch release.

Impact

This is a compatible patch. Default detailed JSON remains unchanged. Agent-facing setup may opt into concise output with one non-duplicated attention list and portable proposal paths.

Deterministic validation

• npm run release:check — passed
• 221 tests — passed
• packed smoke — passed (ultimate-agent-stack@0.9.1, 116 installed files)
• fixture baseline proposal — exact match
• Markdown lint — 0 errors
• git diff --check — passed
• behavior surface: sha256:334b0ac2ae4a05744e757e7b0e2b03fb33441ce8010d4be04105fb54d00dff13

Live evidence boundary

Fresh exact-revision representative harness smoke is required before this draft becomes merge-ready because the version-bound installed bytes changed. No broader full-catalog or provider test is claimed.

[coderabbitai]

Warning

Review limit reached

You’ve reached a temporary PR review limit under our Fair Usage Limits Policy.

Your recent review volume is higher than typical usage, so adaptive limits are currently applied.

Next review available in: 18 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: Path: .coderabbit.yaml

Review profile: ASSERTIVE

Plan: Pro Plus

Run ID: 8516f5f3-ab79-4593-9624-dd615a67bd12

📥 Commits

Reviewing files that changed from the base of the PR and between cf1d756 and 9bb89ed.

⛔ Files ignored due to path filters (1)

• package-lock.json is excluded by !**/package-lock.json

📒 Files selected for processing (6)

• .codex-plugin/plugin.json
• CHANGELOG.md
• bin/ultimate-agent-stack.mjs
• evals/fixture-baselines.json
• package.json
• test/agent-stack.test.mjs

---

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

[samtay32]

Exact-head live compatibility receipt

Candidate commit: 9bb89ed7b5f3b96608c969b2f939d2d487f23a36
Behavior surface: sha256:334b0ac2ae4a05744e757e7b0e2b03fb33441ce8010d4be04105fb54d00dff13
Package: ultimate-agent-stack@0.9.1

Harnesses:

• Codex CLI 0.145.0, configured model gpt-5.6-sol (xhigh for explanation/discovery; low for PRD/direct)
• Claude Code 2.1.89, claude-sonnet-4-6
• Node.js 22.23.2 supplied to the live fixture sessions
• Isolated fixture evidence retained at /tmp/uas-091-live.IjOEpN

Scenario	Codex	Claude Code
explanation only	PASS — explanation, zero writes, zero skill receipts	PASS — explanation, zero writes, zero skill receipts
vague greenfield idea	PASS — DISCOVER draft, exactly one consequential question, no product code; controller + brief receipts recorded	BEHAVIOR PASS / RECEIPT PARTIAL — DISCOVER draft and one question; controller receipt recorded, nested brief receipt missing
complete outside PRD	PASS — EXTERNAL DRAFT, no question, no product code; brief receipt recorded	BEHAVIOR PASS / RECEIPT FAIL — correct EXTERNAL DRAFT and no question, but no activation was persisted
bounded direct task with supporting note	BEHAVIOR PASS / REVIEW LIMITATION — DIRECT route, exact source/test, tests and full local gate pass, controller receipt recorded; normal Codex workspace sandbox blocked a local commit, so exact-commit independent review was not completed	IMPLEMENTATION PASS / EVIDENCE FAIL — exact source/test created and tests pass, but no activation or draft-PR evidence package was persisted

Measured activation receipt rate for the four expected route activations per harness:

• Codex: 4/4
• Claude Code: 1/4
• Combined: 5/8

Deterministic candidate checks remain green: 221 tests, packed smoke, fixture baselines, markdown lint, and diff check.

Release interpretation: the 0.9.1 patch changes concise init/upgrade output and portable proposal-path handling; the tested functional changes pass. The live run also confirms that cross-model compliance with agent-written activation receipts is not reliable enough to claim enforcement. That limitation is disclosed rather than hidden; it is not a regression introduced by this patch. No external provider, deployment, merge, or npm publication occurred during the smoke run.

[qodo-code-review]

PR Summary by Qodo

Release ultimate-agent-stack v0.9.1

⚙️ Configuration changes 📝 Documentation 🧪 Tests 🕐 20-40 Minutes

AI Description

• Bump published/package-facing versions from 0.9.0 to 0.9.1 across entrypoints.
• Move completed “concise install/upgrade” notes into the 0.9.1 changelog section.
• Regenerate deterministic fixture baselines and update version-bound tests for 0.9.1 paths.

Diagram

graph TD
  A["package.json"] --> B["package-lock.json"] --> C["bin/ultimate-agent-stack.mjs"] --> D[".codex-plugin/plugin.json"] --> E["test/agent-stack.test.mjs"] --> F["CHANGELOG.md"]
  B --> G[("Fixture baselines")]

Loading

High-Level Assessment

The following are alternative approaches to this PR:

1. Automated version propagation (postversion/release script)

• ➕ Eliminates manual drift across package.json, lockfile, plugin manifest, and CLI fallback
• ➕ Makes future patch releases faster and less error-prone
• ➖ Adds release-script complexity and requires CI/release discipline
• ➖ Still needs careful handling of regenerated fixtures and test expectations

2. Derive CLI/plugin version dynamically from package.json at runtime

• ➕ Single source-of-truth for version (package.json)
• ➕ Removes need for hardcoded fallback version updates
• ➖ Doesn’t help when package.json truly isn’t available (the reason the fallback exists)
• ➖ Plugin packaging may still require an explicit manifest version

Recommendation: The current approach (explicitly synchronizing versions and regenerating deterministic baselines) is appropriate for a patch release where published artifacts must match exact bytes. Consider adding a small release script to propagate the version to plugin.json, the CLI fallback constant, and version-bound tests to reduce future manual coordination risk.

Files changed (7) +61 / -52

Tests (2) +47 / -47

fixture-baselines.jsonRegenerate deterministic fixture baselines for new installed bytes +46/-46

Regenerate deterministic fixture baselines for new installed bytes

• Updates stored git_head and project_tree_sha256 values for multiple eval scenarios, reflecting the exact revision and tree hash for the 0.9.1 release artifacts.

evals/fixture-baselines.json

agent-stack.test.mjsUpdate version-bound proposal path assertion to 0.9.1 +1/-1

Update version-bound proposal path assertion to 0.9.1

• Adjusts the expected upgrade proposal path to include the 0.9.1 directory, keeping the test aligned with the new release versioning behavior.

test/agent-stack.test.mjs

Documentation (1) +9 / -0

CHANGELOG.mdAdd 0.9.1 release section and move completed notes from Unreleased +9/-0

Add 0.9.1 release section and move completed notes from Unreleased

• Introduces a new 0.9.1 section dated 2026-08-01 and records fixes/behavior shipped in this patch. Keeps 0.9.0 notes intact below.

CHANGELOG.md

Other (4) +5 / -5

plugin.jsonBump Codex plugin manifest version to 0.9.1 +1/-1

Bump Codex plugin manifest version to 0.9.1

• Updates the plugin manifest version from 0.9.0 to 0.9.1 to match the release being published.

.codex-plugin/plugin.json

ultimate-agent-stack.mjsUpdate CLI fallback version constant to 0.9.1 +1/-1

Update CLI fallback version constant to 0.9.1

• Adjusts the embedded package metadata fallback to report version 0.9.1 when package.json is not present/accessible.

bin/ultimate-agent-stack.mjs

package-lock.jsonSynchronize lockfile version metadata to 0.9.1 +2/-2

Synchronize lockfile version metadata to 0.9.1

• Bumps the top-level and root package entries from 0.9.0 to 0.9.1 to keep npm metadata consistent for the release.

package-lock.json

package.jsonBump npm package version to 0.9.1 +1/-1

Bump npm package version to 0.9.1

• Updates the published package version from 0.9.0 to 0.9.1 for the patch release.

package.json

[qodo-code-review]

Code Review by Qodo

🐞 Bugs (0) 📘 Rule violations (0) 📎 Requirement gaps (0)

Great, no issues found!

Qodo reviewed your code and found no material issues that require review

_{To customize comments, go to the Qodo configuration screen, or learn more in the docs.}