skills.rest
diazMelgarejo
diazMelgarejoCOMMUNITY
Explore all skills by this author →
Website𝕏@Melgarejo_yDiazGitHub

pt-orama-lesson-mining

Graduates durable session insights into a local semantic memory store.

Updated Mar 22, 2026
Search Tags:#agent workflow#semantic memory#hermes#knowledge persistence#session insights#lesson mining
One-click install
npx skills add https://github.com/diazMelgarejo/orama-system --skill pt-orama-lesson-mining-diazmelgarejo
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: pt-orama-lesson-mining
Source: https://github.com/diazMelgarejo/orama-system/tree/main/bin/orama-system/skills/hermes-harness/commands/pt-orama-lesson-mining
Command: npx skills add https://github.com/diazMelgarejo/orama-system --skill pt-orama-lesson-mining-diazmelgarejo

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Long agent sessions produce reusable insights that are lost when the session ends. This optional Hermes command captures durable lessons and graduates them into a local semantic memory store so they can be recalled in future sessions. ## Core Features & Use Cases - Lesson graduation: Confirms an insight is durable and operator-approved, then invokes a configured local graduation tool (a harness-local learn CLI, MCP memory server, or manual LESSONS.md edit). - Structured fallback: When no graduation tool is configured, returns a structured output summary for manual paste into memory instead of failing. - Strict boundaries: Never commits secrets or raw harness home exports, and never assumes a specific memory product or repository layout. - Use Case: After a debugging session reveals a reusable fix pattern, an operator invokes the command with a one-line summary and confidence score to persist the lesson for future Hermes sessions. ## Quick Start Ask the agent to run pt-orama-lesson-mining with a one-line summary of the durable insight and a confidence score to graduate it into your local memory store.

Frequently Asked Questions about pt-orama-lesson-mining

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

FAQPage Schema
How do I save a reusable lesson from an agent session?▼

Invoke pt-orama-lesson-mining with a one-line summary of the durable insight and a confidence score. If a local graduation tool is configured, the command invokes it; otherwise it returns a structured summary you can paste into memory manually.

Is Perpetua-Tools required for the Hermes harness to work?▼

No. This command is an optional extension and Perpetua-Tools is not a dependency. Hermes harness authority, bootstrap, council, review, and delegate flows all operate without it.

What memory tools does lesson mining support?▼

It supports any harness-local learn CLI, MCP memory server, or manual LESSONS.md edit configured via args.tool or environment. It does not assume any specific memory product, graduation CLI, or repository layout.

What happens if no graduation tool is configured?▼

The command returns a needs_input status with setup steps, or produces a structured output.summary containing the lesson title for manual pasting into your memory store.

When should I not graduate a lesson?▼

Skip graduation for one-off errors or insights that will not save at least five minutes next time. Only durable, operator-approved insights should be persisted, and secrets or raw harness home exports must never be committed.

Related Skills

Explore More Software Engineering Skills→
zhiyuan-zhang0206zhiyuan-zhang0206
1

ava-use-claude-code-and-codex

Supervises Claude Code and Codex CLI coding agents through file-driven persistent shell sessions.

Community
Advanced
t6adevt6adev
3

self-improvement

Record corrections as structured lessons in tasks/lessons.md and apply them at session start.

Community
Intermediate
project-axproject-ax
1

claude-memory

Read and write structured project knowledge in `.claude/memory/` across sessions.

Official
Advanced
joySUSYjoySUSY

Code Formatting

Format Rust and TypeScript code using justfile tasks after file edits.

Community
Basic