agentleFS
arthur-plautzarthur-plautz/research-toolkit/.claude/skills/project-knowledge/SKILL.md

project-knowledge

Skill0 starsChanged 4 months agoMITView on GitHub

Auto-triggered before any non-trivial implementation task, architectural question, or skill invocation. Loads and synthesizes the project knowledge docs so Claude has accurate project context before proceeding.

16 other projects use a version of this.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

5 rules. 0 are common, 4 are distinctive, and 1 is very specific to this project.

  • 0

    Common

    Most of the 15 edited versions we read keep these.

  • 4

    Distinctive

    This author's own choices. Worth a look.

  • 1

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Before invoking /proposals or /analysis

    line 12
  2. Distinctive:

    When a user question implies uncertainty about how the project is organized

    line 13
  3. Distinctive:

    Do not trigger it for trivial one-line edits, doc fixes, or questions that are clearly self-contained.

    line 15
  4. Distinctive:

    Read all four knowledge docs before proceeding with the task:

    line 19
  5. Distinctive:

    After reading, hold the following as active context:

    line 28

“Common” is measured: kept by at least 6 in 10 of the 15 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Trigger

This skill activates automatically — it is not a user slash command. Trigger it when:

  • Starting any non-trivial implementation task (new feature, refactor, bug investigation)
  • Answering questions about project structure, source layout, or research method
  • Before invoking /proposals or /analysis
  • When a user question implies uncertainty about how the project is organized

Do not trigger it for trivial one-line edits, doc fixes, or questions that are clearly self-contained.

What to load

Read all four knowledge docs before proceeding with the task:

  1. docs/project/overview.md — repository index, components table, mode/variant switches, quick usage context
  2. docs/project/architecture.md — class hierarchy, component summaries, data flow diagram
  3. docs/project/analysis.md — experiment types, data output formats, key metrics, analysis conventions
  4. docs/project/research.md — the currently active research question and scalability rationale

Synthesis

After reading, hold the following as active context:

From overview:

  • Which components/experiment types exist and their variants
  • Any mode/strategy switches and the classes that implement them

From architecture:

  • The main ownership chain between components (who constructs/owns whom)
  • How optional subsystems (if any) wire into the main pipeline
  • Which execution modes are distinct from the main run loop (e.g. a standalone evaluator)
  • Where source lives and how it's organized

From analysis:

  • The primary comparison metric and any population/diagnostic health signals
  • Any fixed axis/unit conventions for plots
  • Any ordering convention that must never be inverted

From research:

  • The current research question and why the compared conditions/modes were chosen
  • Any scaling or generalization hypothesis being tested

After loading

Proceed with the task using this context. Do not summarize the docs back to the user unless they asked for an explanation — load silently and act.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.