agentleFS
dsviddsvid/claude-skills/skills/mapping/visualiser/SKILL.md

visualiser

Skill0 starsChanged 4 months agoMITView on GitHub

Render a compiled map as diagrams a human can read at a glance — constellation of subjects and relations, architecture, coverage-and-resolution, or status — with inferred links drawn differently from sourced ones and unmapped areas drawn as blank. Use when a map or set of findings has become too large to parse by reading, when asked to visualise, diagram or show the map, or when deciding where to survey next.

1 other project uses a version of this.

Adopted by

Most recently taken up by

  • Other versions1 project uses a version of this.

    Word for word · 0

    None found.

    Edited · 0

    None found.

    Wrote their own version · 1

    1. 8ix/neurodiverse-claude-skills★ 6same name

    How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.

  • In this repository4 other files it gives its agents.

The file

We split this skill into the rules it gives an agent

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

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈5

    Distinctive · estimate

    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. Distinctive:

    If the map declares a format version other than v1, stop and say so.

    line 26
  2. Distinctive:

    Put the key on the diagram itself.

    line 51
  3. Distinctive:

    Every diagram carries, visibly: the direction it serves, the date, and a

    line 58
  4. Very specific:

    Never let the prose assert anything the index does not contain.

    line 75
  5. Distinctive:

    Default to mermaid in markdown — it renders in most viewers and diffs in git.

    line 81
  6. Distinctive:

    Keep diagrams readable rather than complete.

    line 88

≈ “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.

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

Visualiser

Speaks map format v1. Reads the index. Renders for a human.

Writes no claims and changes no status. If you find something wrong with the map while rendering it, say so — do not fix it here; that is cartographer.

Why this exists

A map that has to be consulted through an assistant is not readable. The symptom is a human asking what their own map says. Your output should mean they no longer have to ask.

1. Render from the index, never from the claims

🛑 Read the index. It is the only part of the map that knows what is current. Rendering from the underlying claims will draw superseded findings beautifully and confidently, which is the worst failure available here.

If the map declares a format version other than v1, stop and say so.

2. Ask what the human is looking at it for

One question before drawing: what are you trying to see? The answer picks the view. If they do not know, offer the four:

ViewAnswersShape
ConstellationWhat is here and how does it relate?Node graph of subjects and relations
ArchitectureHow is this thing put together internally?Component or sequence diagram
CoverageWhere is the map thin, and where should I survey next?The territory shaded by resolution
StatusWhat is settled, contested, or stale?Board or table by index status

Coverage is the one people do not think to ask for and most often need — it is the only view that shows the shape of what is not known, and it cannot be expressed as a list without becoming another wall of text.

3. 🛑 A diagram is a claim

Diagrams are far more persuasive than prose, so an unsourced one does more damage. Three rules, without exception:

  1. Draw grounding. [source] relations solid; [inferred] relations dashed; [docs] dotted. A guessed relation must never look verified. Put the key on the diagram itself.
  2. Blank is drawn as blank. An unmapped area gets its boundary drawn and its inside left empty, labelled as unmapped. Omitting it reads as "nothing there", which is a claim nobody made.
  3. Mark contested topics as contested — both sides shown, neither presented as the answer.

Every diagram carries, visibly: the direction it serves, the date, and a key. A diagram found later without those is indistinguishable from a confident guess.

Show resolution wherever it can be shown — opacity, border weight, or an explicit R0–R3 label. An R1 sketch drawn identically to an R3 survey is the fluent unsourced map arriving by the back door.

4. Succinct explanation beside every diagram

Three to five lines, and they must include:

  • what the diagram is claiming, in one sentence;
  • what is missing from it — the blank areas, in words as well as shape;
  • what to do next, if the picture implies something (usually: which thin area to survey).

Never let the prose assert anything the index does not contain. If you want to say it and the map does not support it, that is a finding for cartographer, not a caption.

5. Output

Default to mermaid in markdown — it renders in most viewers and diffs in git.

Offer a published artifact when the map is large, when several views are wanted at once, or when the human wants to come back to it: mermaid renders natively there, it gets a URL, and redeploying the same file updates it in place — which suits a map whose resolution increases over time.

Keep diagrams readable rather than complete. A diagram that shows everything shows nothing — if a view needs more than roughly a dozen nodes, split it by area and say how the pieces join.

🛑 Write only to the map location (or to an artifact). Never edit the domain's own documents or source — do not drop diagrams into a README, a status page or a findings file because they would look good there. Renders scatter more easily than any other output, and a diagram pasted into a document nobody re-renders becomes a confident stale picture the moment the index moves. Offer the human the file; let them place it.

🛑 The subject is a READ-ONLY ARTEFACT — rendering is read-only work. Not a list of banned commands, a property of the subject: never git-write (commit, add, checkout, branch, stash, reset, restore, clean, push), never write files into it by any other route, and never send anything outward on its behalf. The repo you are drawing is often one the human does not own; committing a diagram into it puts your output in someone else's history, one push from being public and attributed to them.

6. Report what you could not draw

Finish by naming, in a line or two:

  • topics the index marks contested — they resist a single picture, which is the point;
  • live claims whose last_checked is old, because a diagram makes a stale number look freshly true;
  • 🛑 entries the index marks unverified, or any figure it gives without an n — the same hazard as staleness and a worse one, because a stale number was at least true once. Either draw it with the unverified marking the index gave it, or leave it out and say you did. Never render it as a bare number; a diagram strips every qualifier the index was careful to attach;
  • anything in the map that could not be rendered honestly, and why.

Do not fix the map to make the picture nicer. A map that renders awkwardly is telling you something true about the territory.

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.