wikiimport Agent Skill

Import a single external resource (URL, file, paper, PDF, transcript, meeting note, internal note, paste) into the wiki with a…

Context cost
2.6k tokensestimated from the bundle, loaded when it triggers
Bundle
1 file · 10.4 kBtext throughout, nothing executable
Licence
MITthe author’s own terms
Last change
—v1.3.2
Servers it uses
Noneruns standalone

What it does

Import a single external resource (URL, file, paper, PDF, transcript, meeting note, internal note, paste) into the wiki with a user-reviewable triage step before any wiki page is written. Use when the user points at such a resource and asks to ingest, import, integrate, digest, absorb, review-before-adding, or propose-then-add it into their wiki; when they want a triage step on a single source rather than a straight ingest; or whenever a named source should be brought in with a propose-then-act front end.

What this skill changes about the agent has not been set out here as a list. Its SKILL.md, shown below, says it in the author’s own words.

content

The skill itself

A skill is instructions the model reads, so you can see what you are getting: below is the start of its SKILL.md. The front matter loads with every session, the body loads when the skill triggers, and the whole file comes with the install.

SKILL.md10.4 kB · preview
--- name: wiki_import version: 1.3.2 author: Andreas F. Hoffmann license: MIT ---
9# wiki_import
10
11<wiki_import>
12 <role>Resource-to-wiki reconciliation editor. Capture the resource the user pointed to as raw source, mine the captured raw for durable knowledge, diff against the wiki, and triage the result before any wiki-page write.</role>
13 <orient_first_top>**Read $WIKI/SCHEMA.md once at the start of any session that activates this skill.** The schema declares the domain, page-type enum, tag taxonomy, and conventions every classification and write must honor. The full orientation pass (SCHEMA + index + last ~350 lines of log) is covered by <orient_first> in <policy> below; this top-line note exists so the SCHEMA read is never skipped on a "quick" import.</orient_first_top>
14 <objective>Pull a single named external resource into the wiki end to end: persist the raw, classify each candidate by page type and relation to existing content, surface conflicts between the resource and existing wiki pages, and propose reconciliations the user can accept page by page. Defer all wiki structure, discovery, orientation, raw capture, ingest, and lint behavior to the wiki skill.</objective>
15 <policy>
16 <resolve_skill_bundle>Resolve $WIKI_SKILL before the first bundled tool call and before $WIKI. It is the installed wiki skill bundle holding the scripts and reference docs this skill invokes. The wiki skill's <resolve_wiki_skill_bundle> block is the canonical resolution order, so follow it rather than guessing a path. From this skill's own directory the sibling hub is ../wiki. Every $WIKI_SKILL/... path below resolves through that block.</resolve_skill_bundle>
17 <resolve_first>Resolve $WIKI through the wiki skill's discovery flow before reading the resource. Honor exit-2 ambiguity by presenting candidates and asking the user.</resolve_first>
18 <orient_first>Read SCHEMA.md, index.md, and roughly the last 350 lines of log.md before diffing, so the diff runs against an understood corpus and domain.</orient_first>
19 <confirm_resource>Confirm the resource pointer (URL, file path, paste, PDF, transcript, meeting note, internal note) and its kind with the user before fetching. For a URL to an externally-published article, confirm the target slug under raw/articles/; for a PDF or paper, under raw/papers/; for a meeting note, interview, or spoken-word transcript, under raw/meetings/; for an internal memo, discussion writeup, ad-hoc observation, or internal doc not published externally, under raw/notes/; for a paste, the appropriate raw/ subdirectory by kind. For edge cases (article that embeds a transcript, transcript of a private meeting, paste of unknown provenance, file that fits two buckets equally), consult $WIKI_SKILL/references/raw_taxonomy.md, the canonical reference for bucket meanings and classification heuristics.</confirm_resource>
20 <capture_raw>Route the resource through the wiki skill's Ingest §1, fetching and converting (URL), extracting (PDF), or filing (paste, meeting, note), into raw/<kind>/<slug>.md with the required frontmatter (ingested, body-only sha256, plus source_url for an externally-published source or a relative in-repo source_path; a local file outside the repo takes no path, so excerpt it into the body and note its locality). The ### raw/ Frontmatter contract in $WIKI_SKILL/references/template_schema.md is canonical for these field meanings and for how a mislabeled or legacy sidecar reconciles onto them, so cite it rather than restating the moves. Write the sha256 with python3 "$WIKI_SKILL/scripts/compute_sha256.py" raw/<kind>/<slug>.md. Never invent the value by hand. On re-ingest of the same source, follow the hub's **Re-ingest compares before it writes** rule in <capture_raw_source>, which owns this ordering: write the freshly fetched or converted body to a session-local temporary file outside the wiki tree, under a verbatim copy of the existing sidecar's frontmatter through its closing --- with the recorded sha256 carried over untouched (reproducing the blank line the sidecar keeps beneath that closing ---, since the hash covers everything after it), then run python3 "$WIKI_SKILL/scripts/compute_sha256.py" --check <temp-sidecar> on that temporary file in report-only mode, which writes nothing, while the sidecar on disk stays as it is. ok and exit 0 mean the source is unchanged: skip, leaving the sidecar untouched, with no body rewrite, no hash refresh, and no log.md entry. update and exit 1 mean drift: rewrite the sidecar body to what the source now says, refresh the recorded hash with the write-mode command above, and report the drift both to the user and in this import's log.md entry.</capture_raw>
21 <mine_resource>Read the captured raw end to end. Extract durable claims, decisions, definitions, conventions, comparisons, workflows, and named entities. Skip passing mentions, minor details, and material outside the wiki's stated domain in SCHEMA.md.</mine_resource>
22 <classify>For each candidate, pick a page type from the wiki skill's enum and tag it NEW, EXTEND, CONFIRM, or CONFLICT against $WIKI. Honor the page-threshold rules from the wiki skill: a passing mention does not earn a page.</classify>
23 <surface_contradictions>For every CONFLICT, capture both excerpts verbatim, name the disagreement dimension (factual / definitional / scope / recency / source-quality), and offer two or three concrete reconciliation paths.</surface_contradictions>
24 <propose_then_act>Emit the proposal and wait for user selections before any wiki-page write. The raw capture in step 4 stays; only the entity, concept, comparison, summary, query, and procedure writes wait.</propose_then_act>
25 <defer_writes>Route approved NEW and EXTEND items through the wiki skill's Ingest flow (steps 3 to 6). Route approved CONFLICT items through its contested-page protocol (contested: true, contradictions: frontmatter, both positions recorded with dates and sources).</defer_writes>
26 <single_resource>Source material for the proposal is the captured raw only. Skip claims the resource did not establish, even when they are true in the broader domain.</single_resource>
27 <too_large_to_read_in_one_shot>When Read fails with File content (N tokens) exceeds maximum allowed tokens (25000), do not retry the same call, because the same call fails the same way. Switch strategy: run Bash wc -l <path> to size the file, then Read with offset and limit to walk it in chunks, or run Bash grep -n <pattern> <path> to find the section you need and Read only that range. This applies in particular to Confluence-page tool-result tempfiles under .../tool-results/mcp-claude_ai_Atlassian-getConfluencePage-*.txt and to long transcripts or dense papers, sources that routinely exceed the 25k-token Read limit. The failure mode to avoid is the same-call retry loop of three to five attempts the model defaults to on this error.</too_large_to_read_in_one_shot>
This is a preview of SKILL.md. Download or install the skill to get the whole thing.
In the file
SKILL.md1,397 words
Files1
LicenceMIT
Why you can read it

Nothing in a skill executes. The client loads the text and the model follows it, so a skill can be audited the way a runbook is — by reading it.

What it costs in context

Skills are not billed by the call. They are paid for in context: every token the instructions occupy is a token your code, your diff and your conversation cannot use. Here is what this one takes and when it takes it.

≈150
always loaded
The name and description, so the model knows the skill exists and when to reach for it.
2,450
on trigger
The instruction body, read only when the skill fires.
1.3%
of a 200k window
Ten skills this size would take about 13% of the window before you open a file.
050k100k150k200k context window

2.6k tokens, estimated from the bundle at four bytes to the token, held for the rest of the session once it triggers. Middling. Fine to keep on in a project where you use it weekly, worth unloading in one where you never do.

Servers bill, skills cost

A server charges by the month. A skill charges once per session, in context, and then keeps charging it for as long as the session lives.

Before and after

The same question, put to the same model twice: once as it comes, and once with these instructions loaded.

No worked example has been published for this skill yet.

The procedure it runs

The procedure has not been set out here step by step. It is in the body of its SKILL.md: the start of it is shown above, and the whole file comes with the install.

Prose, not code

These steps are written for a model to follow, not executed by a runtime. It can still be told to skip one, and it will say so when it does.

Servers it uses

None. This skill calls no MCP servers at all.

Everything it needs is in the instructions, so it works in a project with nothing connected — the model reads the file and changes how it works with what it can already reach.

Its SKILL.md does not declare which tools it asks the agent for, so what it touches is whatever the instructions themselves ask for — read them before you install it.

What it asks for
Writes filesnot declared
Network accessnot declared

This SKILL.md carries no allowed-tools line, so neither is declared. A skill grants no permissions of its own — it can only ask for tools your client already has.

What it will not do

Every skill is narrow, and the useful ones say where they stop. These are the jobs this one is the wrong tool for.

What this skill is not for has not been published here. Nothing is implied by that: it is a section the author has not filled in.

What is in the bundle

1 file, 10.4 kB on disk. A bundle is text throughout: the instructions the model reads, plus the templates it fills in.

  • SKILL.md10.4 kB
What is not in it

No dependencies and nothing executable: a skill is text the agent reads, so the bundle is 1 file you can review in full before installing. The MIT licence covers the templates and examples as well as the instructions.

Install

Installing copies the bundle into your project. Nothing runs at install time — the files sit on disk until the model reads them.

# wikiimport · 2.6k tokens when loaded · macOS, Linux mkdir -p .claude/skills && t=$(mktemp -d) && curl -fsSL -o "$t/bundle.tar.gz" https://mcprush.com/api/skills/wikiimport/bundle.tar.gz && tar -xzf "$t/bundle.tar.gz" -C "$t" && [ -d "$t/wikiimport" ] && { [ ! -e .claude/skills/wikiimport ] || { mv .claude/skills/wikiimport "$t/.previous" && echo "Previous copy moved to $t/.previous"; }; } && mv "$t/wikiimport" .claude/skills/ && echo "Installed .claude/skills/wikiimport"
On Windows, or any OS with Node 20+
npx mcprush@latest skill add theafh/wikiimport

Windows PowerShell says 'running scripts is disabled on this system'? Run Set-ExecutionPolicy -Scope CurrentUser RemoteSigned once, or type npx.cmd instead of npx (codex.cmd, gemini.cmd, grok.cmd likewise). PowerShell 7 needs neither. Writes to .claude/skills/wikiimport/ in the current project. For every project instead, run mkdir -p ~/.claude/skills && t=$(mktemp -d) && curl -fsSL -o "$t/bundle.tar.gz" https://mcprush.com/api/skills/wikiimport/bundle.tar.gz && tar -xzf "$t/bundle.tar.gz" -C "$t" && [ -d "$t/wikiimport" ] && { [ ! -e ~/.claude/skills/wikiimport ] || { mv ~/.claude/skills/wikiimport "$t/.previous" && echo "Previous copy moved to $t/.previous"; }; } && mv "$t/wikiimport" ~/.claude/skills/ && echo "Installed ~/.claude/skills/wikiimport" (macOS, Linux) — or add --global to the npx line. The curl line replaces .claude/skills/wikiimport/ only after the download succeeds, so files left there by another skill of the same name, or by an older release, do not mix in. A copy already there is moved aside, not deleted, and the line prints where. It appears as wiki_import in Gemini CLI, Codex, Copilot and Grok (its own name); Claude Code lists it by the folder name, wikiimport.

Which clients pick it up on their own

A skill is a folder of text. Most of these read a skills folder and pick it up without being told; the rest take the same folder another way — an upload in the app, a parameter on an API call — and then choose it by its description just the same.

Claude Code.claude/skills/
Claude DesktopCustomize › Skills
ChatGPTSkills › Create › Upload
Cursor.cursor/skills/
VS Code.github/skills/
Codex CLI.agents/skills/
Gemini CLI.gemini/skills/
Grok Build.grok/skills/
Microsoft CopilotSkills › Upload a skill
PerplexityComputer › Skills › Upload a skill
DeepSeek Harness.dsh/skills/
Zed.agents/skills/
Devin Desktop.devin/skills/
Windsurf (legacy).windsurf/skills/
OpenAI Agents SDKInstructions or sandbox Skills
HTTP / APIa request parameter
This release
Version1.3.2
Publishedno release date on file
PriceFree
Referencetheafh/wikiimport

Versions

v1.3.2 is what is on the shelf; no release here carries a date. Instructions change more often than APIs do — a skill can be rewritten entirely without anything it depends on moving.

v1.3.2
  • No earlier releases have been published to the marketplace.
Pinning

The install command always fetches the release that is current the day you run it — v1.3.2 today — and cannot ask for an earlier one. The copy on your disk does not change until you run it again.

Reviews

—no reviews yet

Nobody has reviewed this skill yet. The rating is the mean of the reviews written here, so there is none until somebody writes the first.

Sign in to review it

A review is tied to an account so it can be answered, edited and taken down by the person who wrote it.

Sign in
Who can post

Anybody with an account, one review each. It carries the version that was current when you wrote it, and the publisher may reply once.

Publisher
Servers0
Claim this skill