agentleFS
YogirajAYogirajA/Claude-Skills/read-wiki/SKILL.md

read-wiki

Skill1 starsChanged 56 days agoMITView on GitHub

Query the knowledge base before doing work. Use when the user asks "what do we know about X", "get context before I update this", "have we seen this pattern", "what connects A and B", or any question that accumulated project or cross-project knowledge could answer. Also invoke proactively when starting a non-trivial code task in a project that has a wiki: read the relevant pages first before touching files. Strictly read-only: never writes to any wiki file. Triggers on /read-wiki or phrases like "check the wiki", "get context", "what's in the wiki about", "have we documented".

The file

We split this skill into the rules it gives an agent

7 rules. About 0 are common, 3 are distinctive, and 4 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.

  • 4

    Very specific

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

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

    Read ~/.claude/CLAUDE.md and look for the ## Wikis section.

    line 18
  2. Distinctive:

    If a project wiki exists for this folder, read it first.

    line 29
  3. Distinctive:

    GK wiki as fallback: If:

    line 31
  4. Very specific:

    Read index.md of the relevant wiki(s).

    line 41
  5. Distinctive:

    Do not offer to update anything.

    line 49
  6. Very specific:

    If the answer would be worth saving, say: "Want me to save this to the wiki?

    line 50
  7. Very specific:

    If the wiki has no answer at all, say so, and point at /research as the way to go get one.

    line 51

≈ “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

read-wiki: Query the Knowledge Base

Pure read. No writes. Never create, edit, or append to any wiki file during this skill.

Step 1: Find all known wikis

Read ~/.claude/CLAUDE.md and look for the ## Wikis section. It lists every wiki with a path to its index.md and a one-line description of what's in it. (A global SessionStart hook may already have injected the project and GK index maps into context at session start; if they are present, use them rather than re-reading.)

Also check if the current working directory has a ./knowledgebase/index.md: a project wiki wired by write-wiki. If it exists and is not already in the registry, use it anyway.

Step 2: Determine which wikis to read

Project wiki first. Match the current working directory against registry entries. If a project wiki exists for this folder, read it first.

GK wiki as fallback. If:

  • The question is cross-cutting (architecture patterns, methodology, prior engagement experience), OR
  • The project wiki doesn't have a relevant answer,

...also read the General Knowledge index.md whose path is given by the Wikis registry in ~/.claude/CLAUDE.md.

Both wikis can contribute to one answer. Synthesize across them, project content takes priority.

Step 3: Read and synthesize

  1. Read index.md of the relevant wiki(s). Do not read every page: scan the index to find candidates.
  2. Open only the pages that are relevant to the question. Follow [[wikilinks]] if they lead somewhere useful.
  3. Synthesize a direct answer with citations: reference the wiki page (and through it, the raw source).

Format: answer first, citations at the end as a compact list. Keep it tight.

Step 4: Stop

Do not offer to update anything. Do not append to log.md. Do not create new pages. If the answer would be worth saving, say: "Want me to save this to the wiki? Run /write-wiki." If the wiki has no answer at all, say so, and point at /research as the way to go get one.

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.