skills.rest
cameronbracken
cameronbrackenCOMMUNITY
Explore all skills by this author →
GitHub

project-learning

Retrieve, record, and verify reusable project lessons with evidence-backed workflow improvements.

Updated Sep 11, 2026
Search Tags:#knowledge management#verification#continuous improvement#lessons learned#research workflow#project learning#waterology
One-click install
npx skills add https://github.com/cameronbracken/waterology-research --skill project-learning-cameronbracken
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: project-learning
Source: https://github.com/cameronbracken/waterology-research/tree/main/skills/project-learning
Command: npx skills add https://github.com/cameronbracken/waterology-research --skill project-learning-cameronbracken

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Research projects repeatedly lose hard-won knowledge: corrections, failed approaches, and verified fixes disappear between sessions, causing the same mistakes to recur. This Skill captures those lessons as structured, evidence-backed records and retrieves the relevant verified ones before substantial work begins. ## Core Features & Use Cases - Context Retrieval: Run waterology learning context "TASK TERMS" or MCP learning_context before substantial work to load a small set of relevant verified lessons scoped to their stated trigger and project. - Lesson Recording: After a user correction, verified fix, or failed approach, record one actionable lesson with trigger, observed outcome, project-relative evidence files, and applicability tags via waterology learning remember lesson.nt or MCP learning_remember. - Verification & Improvement: Append assessments with waterology learning assess LESSON_ID verified and propose shared workflow changes with waterology learning improve improvement.nt, including a regression-check plan. - Use Case: After discovering that a TORC worker only runs through a tested noninteractive wrapper, record the lesson with evidence from notes/worker-smoke.txt, verify it, and have it automatically surface the next time a worker launch is planned. ## Quick Start Ask the AI to retrieve relevant verified project lessons for your current task using the project-learning skill before starting substantial work.

Frequently Asked Questions about project-learning

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

FAQPage Schema
How do I retrieve relevant project lessons before starting a task?▼

Run `waterology learning context "TASK TERMS"` or the MCP `learning_context` tool before substantial work in an initialized project. It returns a small set of relevant verified lessons that apply only within their stated trigger and project scope.

How do I record a lesson after a correction or failed approach?▼

Use `waterology learning remember lesson.nt` or MCP `learning_remember` with a NestedText file containing the trigger, action, observed outcome, applicability tags, and project-relative evidence files. Keep the reason an approach failed, not just the successful replacement.

Can recorded lessons override a frozen protocol or acceptance criterion?▼

No. Verified lessons cannot change a frozen protocol, compute authority, acceptance criterion, or any higher-priority instruction. They only inform behavior within their stated trigger and project scope.

What makes a lesson verified rather than just proposed?▼

A reviewer must confirm the evidence supports the proposed action, then append an assessment with `waterology learning assess LESSON_ID verified`. An intact file alone is insufficient, and changed lesson text or evidence invalidates prior verification.

What should not be included in a project lesson?▼

Do not copy raw transcripts, credentials, or private source contents into a lesson, and do not infer a universal preference from a single task. Global preferences, shared skills, and runtime permissions stay outside automatic promotion.

Related Skills

Explore More Education & Research Skills→
adobeadobe
1.5k

session-retrospective

Capture session corrections and lessons into .ai/memory/lessons.md.

Official
Advanced
KangOxfordKangOxford

checkstate

Reconstruct experiment narratives from git history and logs into status reports.

Community
Advanced
fulcrumresearchfulcrumresearch
223

regimen-demonstrate

Produces .regimen files documenting setup, actions, outcomes from feature verification runs.

Official
Intermediate
gmelligmelli
1

aget-record-lesson

Capture session lessons as templated markdown documents with classification and indexing.

Community
Intermediate