PUMA Research Vault



PUMA Vault

PUMA 307

Benchmark - Local LLM Evaluation Framework


PUMA Repo


Overview

PUMA — PUMA Understanding and Management with Agents

Can language models manage ICT projects? An empirical benchmark of local LLM agents for issue triage and effort estimation in ICT projects.

PUMA is a research-driven platform that benchmarks autonomous AI agents on practical project management tasks. This vault is the unified knowledge management system for the PUMA project. It integrates six complementary methodological layers into a single coherent workspace.


What is this vault?

LayerSystemPurpose
NavigationMOC (Map of Content)Top-level orientation maps
ProductivityGTD (Getting Things Done)Task and commitment management
OrganisationPARAProjects · Areas · Resources · Archive
KnowledgeZettelkastenAtomic, permanent, linked ideas
NumberingJohnny DecimalUnique IDs for every note and folder
EngineeringSDD / BMADSpec-first development artefacts

Research frameworks integrated: EBSE+SLR/PRISMA · DSR · Grounded Theory · MIT Student Method · RCOIF · EGI · AMI · DRCA · IIPR · CoT · Few-Shot · Zero-Shot CoT · CDD · Agent Prompt Engineering.


Vault Structure (Johnny Decimal)

SectionDescription
00 - MetaTemplates, dashboards, plugin configuration
10 - InboxGTD capture point — fleeting notes and quick capture (process daily)
20 - LiteratureAll source materials — papers, books, datasets, tools
30 - PermanentEvergreen Zettelkasten notes — concepts, methods, frameworks, results
40 - ProjectsActive project work — PUMA chapters, specs, experiments, BMAD agents
50 - AreasOngoing responsibilities — research quality, writing, code, ethics
60 - ResourcesReusable assets — prompts, workflows, checklists, glossary, bibliography
70 - ArchiveCompleted and deprecated material
80 - MOCMaps of Content and master indexes — navigation layer
90 - GTDTasks, reviews, sprint boards, someday/maybe lists

Start Here

  1. Daily workflow90 - GTD/95 Reviews/Daily-Review-Template
  2. Project overview80 - MOC/81 Topic-Maps/MOC-PUMA-Master
  3. Research pipeline80 - MOC/81 Topic-Maps/MOC-Research-Pipeline
  4. Prompts library60 - Resources/61 Prompts/
  5. Glossary60 - Resources/64 Glossary/Glossary-Master
  6. Vault guideVAULT-GUIDE.md

Note Lifecycle

Idea/Source → [10 Inbox] → Process → [20 Literature] or [30 Permanent]
                                             ↓
                                   Referenced in [40 Projects]
                                             ↓
                                   Linked in [80 MOC]

Note Types

PrefixTypeExample
FL-Fleeting noteFL-2026-03-15-LLM-idea
LN-Literature noteLN-Tawosi2022-TAWOS
PN-Permanent notePN-Few-Shot-Prompting
PR-Project notePR-PUMA-Ch1-Introduction
SP-Spec noteSP-Triage-Agent-v1
EX-Experiment noteEX-Llama32-ZeroShot-Triage
PT-Prompt templatePT-Claude-RCOIF-Research
MOC-Map of ContentMOC-LLM-Benchmarks

See 00 - Meta/Plugins-Config/Recommended-Plugins for full setup instructions.

Essential: Dataview · Templater · Tasks · Calendar · Periodic Notes · Git · QuickAdd · Kanban · Excalidraw · Smart Connections · Zotero Integration


  • GitHub Repository: pumacp/PUMA
  • Zotero Library: PUMA group library
  • Datasets: Jira SR (Zenodo DOI: 10.5281/zenodo.5901893) · TAWOS (GitHub: SOLAR-group/TAWOS)

Youtube Playlist https://www.youtube.com/@PUMACapstoneProject

Research Discovery https://discovery.researcher.life/my-library/reading-list/1815730

Gemini (GEM) https://gemini.google.com/gem/1h-rxrzZagTsvX59_CGfaoDHjisJ48cz7?usp=sharing

Perplexity Space https://www.perplexity.ai/spaces/puma-6IpatdqAS_yOxg9j69qvAQ

Research Rabbit https://app.researchrabbit.ai/folder-shares/d8244f17-47f7-4f6c-a589-473876578b54

Google Drive https://drive.google.com/drive/folders/1TKbYhYqLIrq7liAPlSF7ztS2Bv0l7vZS?usp=sharing


Full Documentation

See VAULT-GUIDE.md for the complete reference — methodology details, research frameworks, note types, .claude skills, workflow tutorials, plugin configuration, and the full index of all 386 vault files.


Last updated: April 2026 · License: MIT · Built for the PUMA project