Skip to content
Get packs

Live Research Project Manager

Logs research decisions, experiments, dead ends and claims into an ara/ folder after a coding or research session ends.

Source Download skill
Repository
Orchestra-Research/AI-Research-SKILLs
Stars
14,195
Licence
MIT
Files
4 files

What it does

This skill makes the agent act as a research note-taker that runs only at the end of a session. It scans the conversation for decisions, experiments, failed approaches, claims, tricks and changes of direction, then writes them into a structured ara/ folder of YAML and Markdown files. Each entry is tagged as either user-stated or AI-inferred, so the record shows who decided what and how the project evolved.

Great for

  • Researchers or engineers who want an auditable history of how an experiment or project changed direction
  • Developers running long AI-assisted coding sessions who want decisions and dead ends written down for later
  • Teams experimenting with the Agent Native Research Artifact (ARA) format

Skip it if

  • Quick one-off coding tasks where a session log adds no value
  • Situations where the conversation contains sensitive data that should not be written to project files, since the skill copies extracted content into ara/ files
  • Users expecting it to verify that recorded results are true; it records what the conversation shows and labels inferences as AI-suggested

Try asking

  • “At the end of this session, record the decisions and dead ends from our debugging work into the ara/ folder.”
  • “Wrap up: log the experiments we ran today and tag which choices I confirmed versus which the AI suggested.”
  • “Update the research trace with today's pivot away from the previous architecture.”
Step-by-step guide: how to use Live Research Project ManagerInstall, first run, examples, tips and FAQs

Install in 3 steps

  1. 1

    Download the skill

    The ZIP contains the 4 original files, the MIT licence and an attribution note.

    research-manager.zip
  2. 2

    Add it to your agent

    Claude Code (all projects)

    mkdir -p ~/.claude/skills && unzip research-manager.zip -d ~/.claude/skills/

    Claude Code / Agent SDK (this project only)

    mkdir -p .claude/skills && unzip research-manager.zip -d .claude/skills/

    Claude apps (web/desktop): Settings → Capabilities → Skills → Upload skill, and choose the ZIP.

  3. 3

    Ask for the task

    No special command needed — the agent reads the skill's description and loads it when your request matches. Try one of the prompts above.

Get better results

  • Invoke it only after the task is finished, as the skill instructs, so the ara/ files do not pollute working context.
  • Review the generated ara/ files and correct any provenance tags, since inferred entries are labelled ai-suggested by default.
  • Check that the files written under ara/ do not contain secrets or private data from the conversation before committing them to version control.

What's inside

  • references/event-taxonomy.md
  • references/provenance-tags.md
  • references/session-protocol.md
  • SKILL.md

Redistributed unmodified from Orchestra-Research/AI-Research-SKILLs under the MIT licence. The assessment was drafted with AI from the skill's files on 2026-10-29. Skills can instruct your agent to run code — read SKILL.md and any scripts before installing.