| name | session-lifecycle-automation |
| model | opus |
| description | Designs and configures all three loops of solo-operator session automation — session bookends, background maintenance agents, and a delegation flywheel. Use when setting up a new workspace, recovering from the 44% zero-artifact session leak, or preparing to scale delegation across multiple concurrent agents. |
| category | remember-continue |
| version | 1.0.1 |
| tags | ["session","automation","lifecycle","scheduling"] |
| inputs | [{"name":"workspace_path","type":"string","description":"Root path of the workspace to audit and configure automation for","required":true}] |
| outputs | [{"name":"automation_config","type":"ref","format":"markdown","description":"Configured session bookend commands, scheduled agent definitions, and commission directory with template files"}] |
Session Lifecycle Automation Skill
Philosophy
Analysis of 117 sessions revealed the core problem: 44% of sessions produced zero lines of code. Their insights, decisions, and unresolved questions lived only in the chat transcript — and were lost at session close. A solo operator's only scarce resource is attention. Every session that closes without harvesting its output is a compound leak: the next session re-learns, re-decides, and re-discovers.
The solution is not discipline — discipline without hooks drifts to 0% compliance. The solution is three automation loops that run mechanically regardless of whether the operator remembers to invoke them.
All three loops reinforce each other. Background agents discover stale state → they create commission files → commissions produce findings → session bookends deposit findings to memory → background agents tend that memory. The flywheel spins on its own once initialized.
When to Use
- When setting up a new workspace or onboarding a new project to the Dojo ecosystem
- When noticing re-learning across sessions — spending time at session start rediscovering decisions made in prior sessions
- When background maintenance (supply chain sync, data freshness, memory tending) is being done manually and inconsistently
- When delegation to agents is happening ad hoc (paste-the-context style) rather than from a structured commission directory
- When reviewing productivity and finding that sessions close without clear artifacts
Do not use this skill to run a single wrap-up session — use /wrap-up or compression-ritual for that. This skill is for configuring the infrastructure that makes those tools automatic.
Workflow
Phase 1: Audit Current Automation State
Objective: Understand what is and is not already automated before adding anything.
- Check
~/.claude/settings.json for existing hooks on PostToolUse, PreToolUse, Stop, and UserPromptSubmit events.
- List scheduled tasks (cron-style triggers) already configured for this workspace.
- Check whether a commission directory exists (typically
{workspace}/commissions/ or AgenticStackOrchestration/commissions/).
- Identify the dead zones in the operator's schedule — times when no sessions are active and background agents can safely run without contention.
- Produce an audit summary: Loop 1 configured (Y/N), Loop 2 configured (Y/N + which agents), Loop 3 configured (Y/N + commission count).
Phase 2: Configure Loop 1 — Session Bookends
