Release Aiden Agent 0.55.0

[sambitcreate]

Bumps the package and lockfile version to 0.55.0. Merging into main lets the Release desktop workflow publish v0.55.0.

Ships everything merged since 0.54.0

• Google Antigravity provider over ACP, with reusable ACP harness primitives, UI polish and phone support (feat: Google Antigravity provider over ACP, with reusable ACP harness primitives #371, Antigravity UI/UX polish: review fixes and phone support #373); less repeated work for streamed ACP tool activity (Reduce repeated ACP streamed tool activity work #389)
• Workspace panel: launcher tabs, context inspector and terminal docking; composer context-meter visibility is configurable (feat(appearance): make composer context usage visibility configurable #380, feat(workspace): add launcher tabs, context inspector and terminal docking #381)
• Bots rework: Grok-style single-thread Bots on Pi Durable sessions (Bots rework: Grok-style single-thread bots + Pi Durable sessions #377)
• Simulator devices at T3 parity (Android, 3D/iPhone Duo, SSH hosts, device tabs, power features, phone viewer); Simulator devices and Aiden Live screen sharing on by default (Default Simulator devices and Aiden Live screen sharing on #387, Simulator devices: T3 parity (Android, 3D/iPhone Duo, SSH hosts, device tabs, power features, phone viewer) #391)
• Per-subagent model and reasoning effort, plus per-model effort for custom connections (Per-subagent model and reasoning-effort selection #379, Add per-model effort settings for custom connections #384)
• Scheduled approvals that disclose the full task, and tabbed question composers on desktop, iOS and Android (Fix scheduled approvals and tabbed question composers #369)
• GitHub rate-limit and PR reliability fixes; feature-owned chats stay out of every listing (feat(chats): classify feature-owned chats and keep them out of every listing #375, GitHub rate limits and PR reliability (P0) #378)
• Mobile: chats that need attention sort first on Workspaces home (feat(mobile): sort chats needing attention first on Workspaces home #388); Android Material 3 overhaul (Android: Material 3 overhaul, provider logos, full-screen settings, optimistic loading #374); iOS Live Activity end-state test fix (fix(ios): await ActivityKit's delivered end state in Live Activity tests #372)
• Untitled theme structure and SVG thinking marks on desktop, iOS and Android (Adopt Untitled theme structure (keep Aiden radii and palettes) #361, Replace thinking orbs with SVG activity marks (desktop, iOS, Android) #386)
• Behind flags (off by default): Studio foundation, Create Images core and UI, Design Studio core (feat(studio): Studio Foundation — flags, routes, asset store and canvas kit (default off) #376, feat(images): Create Images headless core — ledger, consent, Pi image port (flag off) #382, feat(images): Create Images CI-1 UI — canvas, consent sheet, run panel, Retry (flag off) #383, feat(design): Design Studio DS-1a headless core — projects, runs, partial + Resume (flag off) #385)

Also marks the workspace panel and context inspector plans as merged in docs/plans/README.md.

Checks

• npm run release:check-consumers passes locally.
• main CI was green on ea511e25c (after feat(mobile): sort chats needing attention first on Workspaces home #388 and Reduce repeated ACP streamed tool activity work #389) before Fix scheduled approvals and tabbed question composers #369 merged.

Flaky-test record

• Main CI on 203a8f208: AidenChatTests.testUploadCleanupRevocationDoesNotDeadlockRemoval() failed after 4.8s. It passed on Simulator devices: T3 parity (Android, 3D/iPhone Duo, SSH hosts, device tabs, power features, phone viewer) #391's head and on a single failed-job rerun. Nothing was changed to hide it.
• PR Fix scheduled approvals and tabbed question composers #369 on 983dda313: environment-devices-agent.spec.ts:108 was flaky (Send message stayed disabled after the textarea fill) and passed on a single rerun. Details are in Fix scheduled approvals and tabbed question composers #369.

🤖 Generated with Claude Code

[pullfrog]

No configured credential can serve this run. Every credential configured for openai/gpt-6.1-sol was rejected or is out of quota, so the agent never ran.

• CODEX_AUTH_JSON from account scope: rejected with 401 — re-authenticate with pullfrog auth codex, or switch this repo to a model you hold an API key for. Re-authenticate →

Model settings → · Ask in Discord →

  ^{｜ Rerun failed job ➔ ｜ View workflow run ｜ via Pullfrog ｜ Using gpt-6.1-sol ｜ 𝕏}

[very-hermes-bot]

Hermes Review Bot

Confidence: 5

Engine: agy/gemini-3.8-flash-high
Review mode: full
Head: e62924c1fb2326b3491adaec0fbd61647ce33147
Generated: 2026-10-09T01:55:23+00:00
Reviews: 1

Summary

Prepares the Aiden Agent 0.55.0 desktop release by incrementing the declared application version and synchronizing plan statuses. Both package.json and package-lock.json are updated from 0.54.0 to 0.55.0, satisfying the SemVer extraction and tag resolution contracts in .github/workflows/release.yml and scripts/prepare-ci-release.mjs. Concurrently, docs/plans/README.md updates the plan index entries for the Workspace Panel Redesign and Chat Context Inspector to reflect their merged status. Maintainers should double-check that main CI remains green on the exact head commit prior to merge so the desktop release workflow can publish tag v0.55.0.

Confidence Score: 5/5

Inspected the complete diff, verified lockfile synchronization with package metadata, validated plan references against existing documentation files, and confirmed previous release tagging conventions.

📁 Important Files Changed

• package.json: Bumps application release version from 0.54.0 to 0.55.0.
• package-lock.json: Synchronizes the top-level package version and the root package entry ("") to 0.55.0.
• docs/plans/README.md: Transitions status entries for the Workspace Panel Redesign and Chat Context Inspector from "Complete locally" to "Merged" with notes referencing PRs feat(appearance): make composer context usage visibility configurable #380 and feat(workspace): add launcher tabs, context inspector and terminal docking #381.

Findings

No findings.

📊 Sequence Diagram

Unchanged or not applicable for this change.

Machine-Readable Findings

[]

---

Last reviewed commit: e62924c1fb23
Reviews (1) · Comment /hermes review to trigger a new review · /hermes review full for full re-review