Analysis updated 2026-05-18
Clean up OpenCode sessions older than a set number of days and vacuum the database to free disk space.
Recover a crashed OpenCode session by truncating its transcript to the last 50 exchanges and restarting fresh.
Back up your entire OpenCode system state to a ZIP file before making configuration changes, with automatic rollback on failure.
Use the interactive TUI dashboard to browse all your OpenCode projects, preview transcripts, and view how much space you have reclaimed over time.
| fariello/ocman | a-bissell/unleash-lite | abhiinnovates/whatsapp-hr-assistant | |
|---|---|---|---|
| Stars | 1 | 1 | 1 |
| Language | Python | Python | Python |
| Setup difficulty | easy | hard | hard |
| Complexity | 2/5 | 4/5 | 3/5 |
| Audience | developer | researcher | developer |
Figures from each repo's GitHub metadata at analysis time.
TUI mode requires the Textual and Rich packages, LLM compaction requires an OpenAI-compatible API key.
ocman, short for OpenCode Manager, is a command-line and terminal-based administration tool for OpenCode, an AI coding agent. If you use OpenCode regularly, its session database can grow large over time, sessions can crash mid-run leaving orphaned data, and long conversation transcripts can bloat the context window. ocman gives you a set of tools to clean all of this up without digging into the database manually. The main things it can do: clean up old sessions by age, remove orphaned database records, vacuum the SQLite database to recover disk space, back up your entire OpenCode system state to a ZIP file, and restore from those backups with an automatic rollback if anything goes wrong partway through. It also keeps an audit log of every cleanup run, showing you how much space and cost you have reclaimed over time. For recovering a crashed or oversized session, you can pick a session interactively or by ID, truncate the transcript to the last N exchanges, and optionally pass it through an AI model via an OpenAI-compatible API to produce a compacted summary that can be used as a fresh starting point for a new session. This is useful when a session grew so long that the AI became less effective due to context limits. Alongside the command-line interface, there is a full interactive terminal dashboard built with the Textual library. It has separate tabs for browsing projects and sessions, previewing transcripts, running the recovery wizard with cost estimates, managing the database, viewing historical activity logs, and editing configuration settings. Installation is via pip from PyPI. The basic CLI requires only Python 3.10 and the OpenCode CLI on your system path. The interactive dashboard additionally needs the Textual and Rich Python packages. A standalone single-file script version is available for zero-dependency use. The README is sparse on license information.
A command-line and terminal dashboard tool for managing OpenCode AI agent sessions: clean old sessions, recover crashed ones, compact bloated transcripts, and back up or restore your OpenCode database.
Mainly Python. The stack also includes Python, Textual, Rich.
Setup difficulty is rated easy, with roughly 5min to a first successful run.
Mainly developer.
This repo across BitVibe Labs
Verify against the repo before relying on details.