Menu
Resume a paused, interrupted, or crashed RePPITS workflow by reconstructing state from tasks. Use when continuing prior workflow work after a break.
Install with Codex or Claude Copy this prompt, paste it into Codex, Claude, or another assistant, and let it review the skill page and install it for you.
Based on SOC occupation classification
Reference only (do NOT invoke as an action): token-reduction conventions (subagent isolation, artifact compaction). Applied throughout the other skills.
Reference only (do NOT invoke as an action): the RICE scoring convention for the goal-loop backlog. Read by the loop/make-plan skills.
Implement a single planned issue, optionally inside a bounded autonomous loop. Use to execute one unit of an approved plan.
Run a continuous goal-loop: work a prioritized RICE-scored backlog toward a standing goal, cycle after cycle, consuming/emitting signals. Use for ongoing autonomous improvement toward a goal; not for a single discrete task (use run).
Run the enabled compliance checklists (any of HIPAA / HITRUST / ISO 27001 / SOC 2) against the diff, separating code-verifiable findings from organizational controls. Always run before committing a sensitive change.
Emit or consume cross-loop signals (observations / frictions / opportunities) in the shared signals/ store. Use to record a finding for other loops or teammates, or to read relevant ones before deciding. PHI must never go in a signal.
| name | resume |
| description | Resume a paused, interrupted, or crashed RePPITS workflow by reconstructing state from tasks. Use when continuing prior workflow work after a break. |
Resume a paused, interrupted, or crashed RePPITS workflow by reconstructing state from Claude Code Tasks.
The full /run flow keeps its state in the conversation context, which is lost on a crash, a stop/pause, or a new session. The durable record is the Claude Code Tasks created during Phase 3 (Plan) and advanced during Implement/Test/Secure. This command rebuilds from that record and continues from the last incomplete step.
plans/ directory.TaskList to get all tasks, their status, owners, and blockedBy edges.plans/<feature>/00-parent-design-doc.md) and any matching Linear parent issue if Linear MCP is available.completed tasks = work already done — do not redo them. Trust the Task record over conversation memory.in_progress tasks = work that was interrupted mid-flight. Re-read the corresponding files/diff to see how far it got before continuing. (A goal-loop stamps metadata.heartbeatAt; an in_progress task with a stale or absent heartbeat was interrupted, not running — reclaim and continue it, never redo.)pending tasks with an empty blockedBy = the next available work, lowest ID first.pending tasks still blockedBy open tasks = not yet startable.afplay /System/Library/Sounds/Glass.aiff &) and ask: "Resume from here?".agents/skills/run/SKILL.md from the matching phase. Re-apply the Phase 0 tier and gate policy that was recorded in the plan.metadata.kind=goal task or a plans/loop/<slug>.md with status: active exists, this is a goal-loop, not a one-shot /run. Resume by following .agents/skills/loop/SKILL.md → Self-healing — do this on every (re)entry (reclaim stale in_progress, idempotent-continue), then run cycles continuously to a real exit condition. Confirm once at resume, then do not pause between cycles — that's the loop's contract, and the Stop hook (bin/loop-rearm.py) keeps it alive across turns.TaskList is empty and no plans/ artifact exists, there is nothing to resume — say so and suggest starting a fresh /run.