relevance-first-information-design
Skill0 starsChanged 6 months agoMITView on GitHub
Design or review any human- or agent-facing information surface so the consumer sees the minimum sufficient information at the right abstraction level for the current task, with clear routes to deeper detail only when relevant. Use for chat, docs, code, APIs, dashboards, charts, websites, profiles, messages, UIs, agent instructions, or other information-bearing interfaces.
In this repository13 other files it gives its agents.
- change-consistencySkill
- engineer-agentic-aiSkill
- format-markdownSkill
- information-representation-designSkill
- living-artifactsSkill
- messaging-frameworkSkill
- python-testingSkill
- reviewable-branch-diffsSkill
- review-ready-codeSkill
- securitySkill
- software-designSkill
- swe-methodSkill
- thoughts-to-artifactSkill
The file
We split this skill into the rules it gives an agent
3 rules. About 0 are common, 3 are distinctive, and 0 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈3
Distinctive · estimate
This author's own choices. Worth a look.
0
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 19
Do not hide information that may be necessary to continue correctly.
- Distinctive: line 21
Keep each layer internally coherent in abstraction level.
- Distinctive: line 42
Look for overload, hidden necessary detail, mixed abstraction levels, duplicated deep content, several audiences forced through one path, …
≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Relevance-First Information Design
Core Principle
Decide which information belongs at the current layer and which information should remain available behind a clear drill-down path.
The consumer should encounter the minimum sufficient information needed for the current action, decision, understanding, or navigation step. Additional detail should remain accessible without competing for attention by default.
This skill is about information selection and disclosure, not about choosing whether visible information should be represented as a table, list, hierarchy, diagram, or prose. Use information-representation-design for that.
Requirements
- Reduce unnecessary cognitive load.
- Do not hide information that may be necessary to continue correctly.
- Separate routing from detail.
- Keep each layer internally coherent in abstraction level.
Layers
- Routing layer: Is this relevant now? Where should the consumer go next?
- Working layer: What is needed to act or understand at this level?
- Deep layer: What evidence, implementation detail, history, or explanation is needed only after drill-down?
Cross-Medium Patterns
- Chat: concise answer first; optional depth behind an artifact, link, or explicit drill-down.
- Messaging: short top-level message; optional detail in a thread, call, attachment, or linked document.
- README/docs: route distinct audiences to the relevant deeper pages instead of mixing all procedures together.
- Website/app: overview first; reveal advanced controls or explanations on demand.
- Dashboard/chart: show the primary signal by default; expose secondary dimensions through hover, drill-down, filters, or detail views.
- Code: keep parent functions/modules at a coherent abstraction level and hide lower-level mechanics behind named calls/interfaces.
- Agent instructions: expose applicability before detailed behavior and retrieve deeper instructions only when relevant.
These examples are illustrative, not exhaustive.
Review Existing Work
Look for overload, hidden necessary detail, mixed abstraction levels, duplicated deep content, several audiences forced through one path, or conclusions buried beneath optional explanation.
Self-Check
- What is the consumer trying to do now?
- What is the minimum sufficient information for that task?
- What detail is competing for attention too early?
- What deeper information may become necessary later?
- Is the route to that information obvious?
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
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.