debrief

Scan agent sessions for friction points and append deduplicated issues to a markdown file.

Updated Aug 23, 2026

Search Tags:#documentation #process improvement #workflow optimization #session analysis #agent debrief #tool overhead

Ready to run

npx skills add https://github.com/Jeqqe/jersal-project-template --skill debrief-jeqqe

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: debrief
Source: https://github.com/Jeqqe/jersal-project-template/tree/main/.pi/skills/debrief
Command: npx skills add https://github.com/Jeqqe/jersal-project-template --skill debrief-jeqqe

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Debrief helps the agent capture what slowed it down or made it less reliable during a task, so repeated inefficiencies don’t keep recurring.

Core Features & Use Cases

Self-scanning for friction: Identifies tool overhead, skill overhead, predictable failures, redundant passes, process awkwardness, and specification gaps from the agent’s own session context.
Duplicate-aware logging: Compares new findings against existing entries to avoid redundant records and keep pain-point tracking actionable.
Actionable memory updates: Appends open items into docs/90-agent-memory/workflow-pain-points.md in a consistent format for future runs to act on.

Quick Start

Use the debrief skill after finishing a non-trivial task to record any detected friction points into docs/90-agent-memory/workflow-pain-points.md.

Frequently Asked Questions about debrief

High-intent search queries and answers about installing and using this skill.

FAQPage Schema

How do I log agent inefficiencies and workflow friction after completing a task?▼

To log agent inefficiencies, use a self-scanning process that reads existing pain-point files, detects friction from the session context, deduplicates against prior entries, and appends open issues in markdown format for future improvement.

What types of workflow friction should I track for agent process improvement?▼

Workflow friction tracking covers redundant tool usage, predictable failures, redundant passes, process awkwardness, skill overhead, and specification gaps that degrade agent reliability, speed, and process clarity during task execution.

How does session analysis prevent duplicate friction entries in pain-point logs?▼

Session analysis prevents duplicate friction entries by comparing newly detected friction points against existing pain-point records before appending, ensuring only open, non-redundant issues are logged for actionable tracking.

Can I automate end-of-task documentation without manually reviewing the session?▼

Yes, end-of-task documentation can be automated through self-scanning. The agent reads its own session context to identify friction points and records them without asking the user, appending open issues to the workflow pain-points file.

Does agent debriefing require any dependencies or external components?▼

Agent debriefing requires no external dependencies or components. It operates as a standalone skill that reads existing markdown pain-point files and self-scans session context to append friction logs.

When should I run a workflow debrief for documentation and process optimization?▼

Run a workflow debrief after finishing any non-trivial task where reliability, speed, or process clarity may have degraded. This captures friction points while session context is fresh for future process optimization.