Agent Sessions (macOS)

Session management for Codex, Claude, OpenCode, Cursor, GitHub Copilot CLI, Pi, Antigravity CLI, Hermes, and OpenClaw on macOS. Search, inspect, save, and resume local AI-coding sessions from CLI tools, desktop apps, and IDE agent surfaces.

• Requires: macOS 14+
• License: MIT
• Security & Privacy: Local-only. No telemetry. Details: docs/PRIVACY.md and docs/security.md

Download Agent Sessions 4.5 (DMG) • All Releases • Install • Resume Workflows • Development

New in 4.5 — See what each session is costing you: the Session Runway can now report every active session's API-equivalent spend per hour, priced per model, alongside 5-hour, weekly, and token rates. And the Quota Meter holds still — it no longer resizes when your pointer crosses it, so you can drag it where you want it. See what's new ↓

Overview

Agent Sessions is a local-first Mac app for finding useful work that coding agents already wrote to disk. It brings Codex, Claude, OpenCode, Cursor Agent, Hermes, OpenClaw, Antigravity, GitHub Copilot CLI, and Pi histories into one searchable view, with transcript inspection, image browsing, saved-session recovery, and resume commands for supported CLIs.

New in 4.0 — Session Runway: live per-session quota burn-rate

Sessions search with transcript and image preview

Saved Sessions with restore actions

Image Browser for visual session outputs

Security & Privacy

• Local-first: session data stays on your Mac.
• No telemetry, analytics, remote logging, advertising identifiers, or session-history uploads.
• Reads local session folders you choose, plus supported default CLI locations.
• Builds local indexes/databases for search and navigation.
• Explicit actions may open Terminal/iTerm2 resume commands or run status/probe cleanup workflows.
• The only network activity is optional Sparkle update checks and an optional read-only fetch of a public model-price list (for the runway's $ estimate) — neither sends any personal or session data.

Details: docs/PRIVACY.md and docs/security.md.

What's New in 4.5

TL;DR - See what each session is costing you: the Session Runway can report every active session's API-equivalent spend per hour, priced per model. And the Quota Meter holds still — it no longer grows when your pointer crosses it, so you can drag it where you want it.

Highlights: The Session Runway now measures whatever you ask it to — 5-hour burn, weekly share, raw tokens/hour, or dollars. The $ rate prices each model in a session at its own rate, so an orchestrator on Opus driving subagents on Sonnet is costed at what each actually runs at rather than blended into one number. Meanwhile the Quota Meter stops fighting you: it used to expand whenever the pointer crossed it, moving the window while you were aiming at it. Now the pointer only moves it and right-click summons its controls. The cockpit also gets one View-menu entry for the whole choice — Quota Meter, Compact, Full, or Off — so the pinned window can finally be closed.

New in 4.5:

• Dollar burn — each active session's API-equivalent cost per hour, priced per model, not blended.
• Selectable runway rates — 5-Hour, Weekly, Tokens, or Dollars, chosen from the meter and remembered.
• A Quota Meter that stays put — no pointer-driven resizing; right-click for the toolbar; drag it anywhere.
• One View-menu entry — Quota Meter / Compact / Full / Off, with ⌘⌥⇧C and ⇧⌘M. ⌘W closes the cockpit.

Previous release — 4.4: Codex usage keeps working when OpenAI pauses the 5-hour limit. Full history in the changelog.

What's New in 4.4

TL;DR - Codex usage keeps working when OpenAI pauses the 5-hour limit. Agent Sessions detects the change, shows a calm "no limit" for 5h while keeping your weekly usage accurate, and the Session Runway shows honest per-session token throughput instead of a misleading rate.

Highlights: OpenAI temporarily removed Codex's 5-hour rate-limit window, which left the Quota Meter mislabeling weekly usage as "5h" with the wrong reset. Agent Sessions now recognizes this automatically — the 5h line reads a calm "no limit", your weekly usage stays accurate, and it snaps back on its own when OpenAI restores the window. With no 5-hour budget to measure against, each active Codex session's runway now shows its real token throughput (e.g. 412K tk/h) instead of a burn rate scaled to the wrong window, and a new format guardrail shows "can't verify" rather than a wrong number if a provider changes its usage data unexpectedly.

New in 4.4:

• Codex 5-hour drop handled — auto-detects when OpenAI pauses the 5h window; shows "no limit" for 5h, keeps weekly accurate, and self-recovers when it returns.
• Honest Session Runway for Codex — active sessions show real token throughput (412K tk/h) while the 5h window is gone; reverts to the familiar 5h "m/h" once it's back.
• Format guardrail — shows "can't verify" instead of a wrong number when a provider's usage format changes unexpectedly.

What's New in 4.3.2

TL;DR - A much quieter Quota Meter. The usage meters no longer re-read and re-parse your entire session history every few seconds while idle, so your Mac stays cool and quiet — plus a more resilient Claude Web API path and a calm state for expired logins.

Highlights: The usage surfaces used to re-parse the whole session corpus on every 5-second refresh; they now cache each file's parse, cutting idle CPU from roughly 25–41% down to about 11% (measured on Release). The "active burn" shimmer pauses when nothing is burning and honors Reduce Motion. On the Claude side, the Web API path is far more resilient — it surfaces a Full Disk Access problem clearly instead of failing silently, recovers from retries instead of stalling, and adds a Test Web API self-check in Preferences. An idle, expired Claude token now shows a calm "no active session" state instead of a misleading error.

New in 4.3.2:

• A much quieter Quota Meter — usage meters cache each file's parse instead of re-reading your whole history every 5 seconds, cutting idle CPU from ~25–41% to ~11%.
• Reduce Motion aware — the "active burn" shimmer runs only while a session is burning and honors the system Reduce Motion setting.
• Resilient Claude Web API — a Full Disk Access problem now shows a clear cause instead of failing silently, retries recover instead of stalling, and a new Test Web API button runs an end-to-end self-check.
• Calm expired state — an idle, expired Claude token shows "no active session" instead of a misleading error.

Previous release — 4.3.1: A rebuilt first run and one-click "Fix" for usage tracking. Full history in the changelog.

Core Features

• Browse and search Codex CLI, Codex Desktop, and Codex VS Code sessions in one place.
• Browse Claude CLI and Claude Desktop sessions with consistent labels and project context.
• Browse Cursor Agent transcripts from Cursor's local storage, enriched with Cursor chat metadata when available.
• Hermes Agent sessions participate in browsing, search, filtering, analytics, and resume workflows, including current ~/.hermes/state.db storage.
• OpenClaw sessions participate in browsing, search, filtering, deleted-session visibility, and resume workflows while ignoring trajectory traces.
• Pi CLI sessions now participate in browsing, search, filtering, and resume workflows.
• Unified browsing across supported agents, with strict filtering, saved sessions, and a single session list.
• Unified Search and Image Browser across sessions, plus in-session Find for fast transcript navigation.
• Readable tool calls/outputs and navigation between prompts, tools, and errors.
• Right-click Copy Resume Command or Resume for supported CLI sessions, with Terminal.app, iTerm2, and Warp launch targets.
• Agent Cockpit is the live command center for active Codex CLI, Claude CLI, and OpenCode CLI iTerm2 sessions, with a compact Quota Meter for always-on Codex and Claude usage visibility, freshness diagnostics, and projected run-out alerts.
• Local-only indexing designed for large histories.

Agent Cockpit (Beta)

Agent Cockpit is the live command center for active iTerm2 Codex CLI, Claude CLI, and OpenCode CLI sessions, with shared active/waiting summaries and live Claude usage tracking.

Quota Meter with Session Runway per-session burn

Agent Cockpit

Agent Cockpit Setup

Prerequisites

• Agent Sessions with Live Sessions enabled
• iTerm2
• Agents running in iTerm2

Ideal Setup

• Set the iTerm window title to the repo name
• Run that repo's agents in that window
• Give each tab/session its own clear name
• Use the same name for the tab, session, and badge

Layout

• One repo per desktop/Space if possible
• Or keep several on one desktop if you prefer
• Keep Agent Cockpit pinned in a corner so you can always see activity
• Click from the cockpit to jump straight to a session

Install

Option A — Download DMG

1. Download AgentSessions-4.5.dmg
2. Drag Agent Sessions.app into Applications.

Option B — Homebrew

brew tap jazzyalex/agent-sessions
brew install --cask agent-sessions

Automatic Updates (Sparkle)

Agent Sessions uses Sparkle for automatic updates (signed + notarized).

To force an update check (for testing):

defaults delete com.triada.AgentSessions SULastCheckTime
open "/Applications/Agent Sessions.app"

Documentation

• Guides:
  • Codex local history: search Codex CLI, Desktop, and VS Code sessions
  • OpenCode SQLite history: browsing old runs
  • Claude Code JSONL history: what you can recover locally
  • Cursor Agent local history: search Cursor Agent transcripts
  • Hermes Agent state database history
  • OpenClaw local agent history
• Release notes: docs/CHANGELOG.md
• Monthly summaries: docs/summaries/
• Privacy: docs/PRIVACY.md
• Security: docs/security.md
• Maintainers: docs/deployment.md

Resume Workflows

• Right-click any supported CLI session and choose Copy Resume Command to get the exact CLI command for that session.
• Open supported Resume sessions in your preferred terminal: Terminal.app, iTerm2, or Warp.
• Use Unified Search (across sessions) and Find (within a session) to jump to relevant tool calls and outputs quickly.

Privacy & Security

• Local-only. No telemetry.
• Reads agent session directories in read-only mode:
  • ~/.codex/sessions
  • ~/.claude/projects
  • ~/.gemini/antigravity/brain
  • ~/.copilot/session-state
  • ~/.cursor/projects and ~/.cursor/chats
  • ~/.factory/sessions and ~/.factory/projects
  • ~/.hermes/state.db and ~/.hermes/sessions
  • ~/.openclaw/agents and legacy ~/.clawdbot/agents
  • ~/.pi/agent/sessions
  • ~/.local/share/opencode/opencode.db and ~/.local/share/opencode/storage/session
• Details: docs/PRIVACY.md and docs/security.md

Development

Prerequisites:

• Xcode (macOS 14+)

Build:

xcodebuild -project AgentSessions.xcodeproj -scheme AgentSessions -configuration Debug -destination 'platform=macOS' build

Tests:

xcodebuild -project AgentSessions.xcodeproj -scheme AgentSessionsTests -destination 'platform=macOS' test

Contributing:

• CONTRIBUTING.md

License

MIT. See LICENSE.