naman10parikh/agentgrid / skills/agentgrid-sessions
agentgrid-sessions
VerifiedSave, restore and replay agentgrid layouts. Save a working grid (layout, pane names, agent commands, conversation ids), restore it after a crash or a reboot, and launch the built-in team presets. Use when the user says "save this grid", "restore my session", "launch a preset" or "set up the same team again".
agentgrid-sessions: save, restore and replay grids
When to use
- A grid is set up the way the user likes, and they want it back tomorrow.
- A terminal closed, a machine rebooted, or an agent crashed, and the team must come back as it was.
- The user wants a ready-made team layout (a preset) instead of building one by hand.
Steps
-
Save the current grid:
agentgrid save my-teamThis records each pane's agent command, directory, name and conversation id in
~/.agentgrid/sessions/my-team.json. -
Restore it later:
agentgrid restore my-team # layout, names, and resumed conversations agentgrid restore my-team --no-start # layout and names only, no agentsRestore resumes each conversation with the agent's own flag: Claude
--resume <id>, Codexcodex resume <id>, Gemini--resume latest, Goosesession resume --last. -
Use a preset.
agentgrid preset listshows them;agentgrid preset show <name>prints one;agentgrid launch <name>starts it. Built-in presets ship inpresets/:Preset Panes What it sets up dev-sprint4 Frontend, Backend, Tests, Docs mixed-agents6 two Claude, two Codex, one Gemini, one scratch shell design-sprint4 Discovery, Composer, Builder, Customer-Zero research-sprint4 a lead researcher, web, repositories, a synthesizer sparc-pipeline6 a lead, then Specification, Pseudocode, Architecture, Refinement, Completion anti-drift-squad8 a coordinator, research, architecture, two coders, tests, review, QA research-swarm9 a lead, seven sources (GitHub, web, docs, forums, X, competitors), a synthesis pane full-company9 a lead, architecture, frontend, backend, QA, content, research, DevOps, design earning-factory9 a lead, six task agents, a dashboard, QA From another coding agent these are MCP tools:
session_save,session_restore,preset_listandpreset_load. -
Re-assign work after a restore with the
agentgrid-orchestrateskill: a restore brings back the grid and the conversations, not the missions in flight.
Output
A saved session file in ~/.agentgrid/sessions/, or a grid rebuilt from one. Report back to the user: the session
or preset name, the number of panes, and whether each agent resumed its conversation.
Do not
- Do not restore into a window that already holds a grid: run
agentgrid killfirst. - Do not treat a saved session as a backup of the agents' work: commit the work to git.
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| reverse-engineer-anythingmorluto/reaReverse engineer native, managed, Electron/JavaScript, packaged, firmware, and browser targets with REA, and build agent environments (AgentEnv bundles) of web applications from a live instance or a capture, and of Linux desktop apps with a task from a request. Use shipped-artifact or requested runtime evidence to explain features, compare versions, decompile code, or guide a reconstruction. Skip REA for ordinary source-repository architecture analysis. | Stars: 469k | Updated | Claude CodeCodex | Verified | Install |
| diagnosing-superpowersobra/superpowersUse when a superpowers session went wrong and your human partner wants to know why — repeated work, ignored plans, stumbles, poor results, a skill that didn't fire, "it took too long", "why is it so expensive", "what is it doing" — or wants to build a bug report for the superpowers maintainers, for the current session or a past one identified by id or path, on any harness. | Stars: 321k | Updated | Claude CodeCodex | Verified | Install |
| dispatching-parallel-agentsobra/superpowersUse when facing 2+ independent tasks that can be worked on without shared state or sequential dependencies | Stars: 321k | Updated | Claude CodeCodex | Verified | Install |
| executing-plansobra/superpowersUse when executing an implementation plan in the current session as the implementer yourself — your human partner chose inline execution, or no subagent tool is available | Stars: 321k | Updated | Claude CodeCodex | Verified | Install |