Agent skill

Wiki Grow

by flonat in flonat/flonat-research

Auto-promote eligible project knowledge articles into Research Vault concepts with generated provenance metadata.

MITAuto-check: notesWriting & Content

Install Wiki Grow

skills CLI
$ npx skills add flonat/flonat-research --skill wiki-grow -a claude-code

Project install by default; add -g for ~/.claude/skills/.

GitHub CLI
$ gh skill install flonat/flonat-research wiki-grow --agent claude-code

Project scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).

Manual copy
$ git clone --depth 1 https://github.com/flonat/flonat-research.git skills-src && mkdir -p .claude/skills && cp -r skills-src/skills/wiki-grow .claude/skills/wiki-grow && rm -rf skills-src

Use ~/.claude/skills/ instead of .claude/skills for a personal install. The folder must contain SKILL.md.

Claude Code skills documentation · loads skills from .claude/skills/

Facts

Skill name
wiki-grow
GitHub stars
146
Token cost
~1.8k tokens
SKILL.md length
718 words
Files
1
Skills in repo
83
Repo updated
First seen
Licence
MIT

At a glance

Auto-promote eligible project knowledge articles into Research Vault concepts with generated provenance metadata.

  • Works in 5 steps: The slug appears as a project knowledge… → The slug is mentioned in ≥1 corpus… → The slug is not project mechanics… → …
  • Running the scheduled knowledge-corpus promotion
  • SKILL.md covers When to Use, When NOT to Use, Architecture and Promotion Thresholds, plus 6 more sections
  • Calls uv

What it does

Wiki Grow is an agent skill from flonat/flonat-research. Auto-promote eligible project knowledge articles into Research Vault concepts with generated provenance metadata. Use when running the scheduled knowledge-corpus promotion or previewing candidates interactively. Not for consolidating overlapping concepts; use $wiki-merge.

Its SKILL.md is about 1.8k tokens, which your agent loads only when the skill is triggered. It is a single SKILL.md file with no bundled scripts.

It sits in Writing & Content, covering Technical writing. The repository describes itself as: Shareable Claude Code + Codex infrastructure for PhD researchers — skills, agents, hooks, and rules for academic workflows. The licence is MIT.

When your agent uses it

  • Running the scheduled knowledge-corpus promotion
  • Previewing candidates interactively

Example prompts

  • “/wiki-grow”

Requirements

  • Python 3
  • Pre-approved tools (allowed-tools): Bash, Read

Workflow steps

5 steps, taken from the first numbered list in SKILL.md.

  1. The slug appears as a project knowledge file in at least one
  2. The slug is mentioned in ≥1 corpus document (atlas topics +
  3. The slug is not project mechanics (PROJECT_MECHANICS_RE: dated
  4. No vault concept already exists at ~/vault/concepts/.md.
  5. The slug is not in ~/vault/concepts/.denylist.

What it can do on your machine

Read from SKILL.md and the folder at commit da27600. It shows what the files ask for, not the result of running them.

  • Tool permissions

    Pre-approves these tools, so the agent can use them without asking each time:

    • Bash
    • Read

    From allowed-tools in the SKILL.md frontmatter.

  • Runs code

    Shell commands in SKILL.md call:

    • uv

    From the folder's file list and the shell code blocks in SKILL.md.

  • Network

    No URLs in SKILL.md. Its commands use uv, which can reach the network depending on how they are called.

    From URLs in SKILL.md, links to its own repository left out.

  • Credentials

    Names no API keys, tokens, secrets or passwords.

    From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.

Context cost

Wiki Grow loads about 1.8k tokens when it runs. Until then it costs about 71 tokens; SKILL.md has 718 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it
~71
When it runs · the whole SKILL.md, loaded when a task matches
~1.8k

Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.

Safety

Auto-check: notes

The automated check noted patterns worth knowing about, such as sudo or a known installer.

  • NotePre-approves every shell command (allowed-tools: Bash)SKILL.md
    allowed-tools: Bash, Read

Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); files beside SKILL.md are not scanned.

SKILL.md

The full file from flonat/flonat-research at commit da27600, republished under its MIT licence (© flonat). 718 words, ~1,806 tokens.

Download SKILL.mdSave it as .claude/skills/wiki-grow/SKILL.md (or your agent's skills folder).
name
wiki-grow
description
Auto-promote eligible project knowledge articles into Research Vault concepts with generated provenance metadata. Use when running the scheduled knowledge-corpus promotion or previewing candidates interactively. Not for consolidating overlapping concepts; use $wiki-merge.
allowed-tools
Bash, Read
argument-hint
--dry-run (default) | --autonomous to actually write

Wiki Grow — Autonomous Concept Promotion

Saturday-cron worker that grows ~/vault/concepts/ from existing project knowledge articles. Designed for zero cognitive load: it surfaces results in the Maintenance Inbox; suppression is a single line in a .denylist file. the user never has to decide "promote this y/n" at the moment of running.

When to Use

  • The Saturday 06:45 cron runs it automatically. Don't invoke manually unless you want to:
    • Preview what would be promoted: wiki-grow --dry-run
    • Force a run after curating the denylist or scaffolding new knowledge: wiki-grow --autonomous

When NOT to Use

  • For per-project knowledge compilation — that's compile-knowledge.
  • For checking which concepts already exist — that's just ~/vault/concepts/ or the wiki at wiki.example.com.
  • For tearing down a concept — delete the file and add the slug to the denylist (one line each).

Architecture

Three components, all in this repo:

FileRole
scripts/wiki-promote-scan.pyInventory scanner. Walks vault concepts + project knowledge folders + atlas/book prose. Outputs /tmp/wiki-promote-scan.{md,json} with corpus mention counts and the wikilink graph.
scripts/wiki-grow.pyOrchestrator. Reads the scan, applies thresholds, copies candidates into ~/vault/concepts/<slug>.md with auto-generated frontmatter and a draft banner. At copy time, [[wikilinks]] whose targets resolve to no vault file (atlas/concepts/venues/people/themes — same semantics as atlas-workspace's dead_wikilinks() validator) are converted to plain text; links between slugs co-promoted in the same run survive. Added 2026-06-12 after verbatim copies created 38 dead links on the wiki.
scripts/weekly-wiki-grow.shLaunchd-callable shell wrapper. Cron entrypoint.
~/Library/LaunchAgents/com.example.weekly-wiki-grow.plistSaturday 06:45 schedule.

Promotion Thresholds

A slug is promoted on a given run if all of the following hold:

  1. The slug appears as a project knowledge file in at least one <project>/knowledge/ directory.
  2. The slug is mentioned in ≥1 corpus document (atlas topics + book chapters + project CLAUDE.md files) — MIN_CORPUS_MENTIONS, env-overridable via WIKI_GROW_MIN_MENTIONS. A knowledge article the curated corpus never references is project-internal, not a general concept. (History: the original gate was ≥3, retired 2026-07-03 as empty for this corpus; briefly removed entirely, which over-promoted 0-mention project notes; the ≥1 middle ground was set 2026-08-07.)
  3. The slug is not project mechanics (PROJECT_MECHANICS_RE: dated notes, -findings/-results/-decisions, hpc-, -roster, r1-, related-literature, etc.) — a secondary exclusion applied regardless of mention count.
  4. No vault concept already exists at ~/vault/concepts/<slug>.md.
  5. The slug is not in ~/vault/concepts/.denylist.

Run cap: at most 8 candidates per run (MAX_CANDIDATES_PER_RUN, env WIKI_GROW_MAX). Excess roll over to the next Saturday. To scope a single manual promotion, set WIKI_GROW_MAX=1 — do NOT denylist the other candidates, which only frees cap slots for backlog items to promote in their place.


Stub Format

Each promoted concept gets:

markdown
---
title: <Concept Name>
slug: <slug>
type: concept
auto_generated: true
source_project: <project-slug>
source_path: /full/path/to/knowledge/<slug>.md
promoted_on: 'YYYY-MM-DD'
corpus_mentions_at_promotion: N
related_topics: []
---

> **Auto-promoted draft.** Seeded from `<project>/knowledge/<slug>.md`
> on YYYY-MM-DD (corpus mentions at promotion: N). Body is verbatim
> from the source — refine for project-agnostic use when read. To
> remove, add the slug to `~/vault/concepts/.denylist`.

<body verbatim from project knowledge article, leading H1 stripped>

The auto_generated: true flag causes the concept overview page on wiki.example.com to render a small amber "auto-promoted draft" badge next to the title — the visual contract that distinguishes machine seeds from human-curated entries.


Show full SKILL.md (271 more words)Show less

Suppression — the Denylist

Single source of truth: ~/vault/concepts/.denylist. One slug per line; # lines are comments.

When a denylisted slug is encountered:

  • Scanner still picks it up (it's part of the corpus).
  • Orchestrator skips it during candidate generation.
  • The slug never gets a vault concept stub, no matter how many corpus mentions accumulate.

To stop the noise from a slug you've decided isn't a concept:

# ~/vault/concepts/.denylist
prompt-sensitivity      # too narrow — covered by llm-as-judge
threat-models           # generic — never promote

If you change your mind, remove the line. The next Saturday run will re-promote on threshold.


Output

Per-run summary in ~/vault/inbox/wiki-grow-YYYY-MM-DD.md:

# Wiki promotion run (YYYY-MM-DD)

- Promoted: **3** new concepts to `~/vault/concepts/`
- Skipped (already in vault): 6
- Skipped (denylist): 2
- Below threshold (3+ mentions): 12

## Promoted slugs
- [`strategic-classification`](https://wiki.example.com/concept/strategic-classification) (14 mentions, from example-project-e)
- ...

The maintenance inbox (atlas.example.com/inbox) surfaces this file as a single line "Wiki grew: N concept(s) auto-promoted" so you have weekly signal without weekly triage work.


Modes

ModeInvocationBehaviour
Dry-run (default)uv run python scripts/wiki-grow.pyPrints the candidate plan, writes nothing. Safe to run ad-hoc.
Autonomousuv run python scripts/wiki-grow.py --autonomousWrites stubs + inbox summary. Used by cron.

--autonomous here matches the global flag convention (--autonomous / -y) — see rules/phased-work.md.


When the LLM Layer Lands

v1 of this skill is deterministic copy + banner — no LLM dispatch per candidate. Project-specific framing in stub bodies is acceptable short-term cost in exchange for fast, idempotent, zero-key cron runs.

When the LLM refinement step is added (v2), it will replace the body of _promote_one() in wiki-grow.py with a sub-agent dispatch that reads the source article and writes a project-agnostic version. The stub frontmatter (auto_generated: true, etc.) stays the same so existing pages don't need migration.


Cross-References

SkillRelationship
compile-knowledgeUpstream — writes the project knowledge articles wiki-grow promotes from. Must run first to seed candidates.
research-atlas-auditChecks vault concept coverage as part of its routine audits.
process-inboxTriages the weekly summary item.

© flonat, MIT. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file

Files

Just SKILL.md in skills/wiki-grow of flonat/flonat-research.

Open the folder on GitHubat commit da27600

Compare with similar skills

Wiki Grow next to the 5 skills that share the most tags, products or categories with it. Stars are the repository's; “used in” counts other GitHub owners with a copy.

Wiki Grow compared with similar skills
SkillStarsUsed inTokensAuto-checkLicenceRepo updated
Wiki Grow this skillflonat/flonat-research146—~1.8kAutomated safety check: NotesMIT
JavaScript Concept Fact Checkerleonardomso/33-js-concepts67k1 repos~5kAutomated safety check: PassMIT
Beads Documentation Style Guidegastownhall/beads28k—~3.2kAutomated safety check: PassMIT
JavaScript Concept Page Workflowleonardomso/33-js-concepts67k—~3.9kAutomated safety check: PassMIT
JS Concept Resource Curatorleonardomso/33-js-concepts67k—~4.9kAutomated safety check: PassMIT
Technical Writing Standardcursor/plugins10k10 repos~2.4kAutomated safety check: PassNone

Similar skills

  • JavaScript Concept Fact Checker

    leonardomso/33-js-concepts

    Verifies the technical accuracy of JavaScript concept pages by checking code examples, MDN and ECMAScript claims and external links through a five-phase method.

    67k GitHub starsUsed in 1 repo~5k tokens
    Writing & ContentAuto-check passed
  • Sets the house style for the beads user docs: the canonical concept model, required terminology, prose and diagram conventions, and checks before docs work is done.

    28k GitHub stars~3.2k tokensUpdated today
    Writing & ContentAuto-check passed
  • JavaScript Concept Page Workflow

    leonardomso/33-js-concepts

    Orchestrates five skills to produce a complete JavaScript concept documentation page, from resource curation through writing, tests, fact-checking and SEO.

    67k GitHub stars~3.9k tokensUpdated 29 days ago
    Writing & ContentAuto-check passed
  • JS Concept Resource Curator

    leonardomso/33-js-concepts

    Finds, vets, writes up and maintains external articles, videos and courses for JavaScript concept pages, including audits for broken and outdated links.

    67k GitHub stars~4.9k tokensUpdated 29 days ago
    Writing & ContentAuto-check passed
  • Official

    Applies four layers of technical-writing rules to docs, RFCs, readmes, PR descriptions and commit messages so a tired engineer follows them on the first read.

    10k GitHub starsUsed in 10 repos~2.4k tokens
    Writing & ContentAuto-check passed
  • Writing Docs

    nukeop/nuclear

    A skill your agent uses when writing or editing documentation in packages/docs.

    19k GitHub stars~907 tokensUpdated 3 days ago
    Writing & ContentAuto-check passed

More from flonat/flonat-research

All 83 skills in this repo
  • Latex Posters

    flonat/flonat-research

    Create a large-format academic poster in LaTeX using beamerposter, tikzposter, or baposter.

    146 GitHub stars~1.5k tokensUpdated 10 days ago
    Auto-check: notes
  • Skill Creator

    flonat/flonat-research

    Create, revise, and evaluate reusable AI workflow skills, including trigger-quality tests.

    146 GitHub stars~4.4k tokensUpdated 10 days ago
    Auto-check passed
  • DOCX

    flonat/flonat-research

    Create, read, edit, or convert Microsoft Word documents while preserving professional document structure.

    146 GitHub stars~1.2k tokensUpdated 10 days ago
    Auto-check passed
  • PDF

    flonat/flonat-research

    Read, create, combine, split, rotate, OCR, watermark, secure, or extract content from PDF files.

    146 GitHub stars~488 tokensUpdated 10 days ago
    Auto-check passed
  • Init Project Orchestration

    flonat/flonat-research

    Create or migrate project-level agents, repeatable project workflows, and planning state from one client-neutral contract, then render repository-scoped adapters for both Claude Code and Codex.

    146 GitHub stars~1.6k tokensUpdated 10 days ago
    Auto-check passed
  • Pre Commit Audit

    flonat/flonat-research

    Deliver a fast pre-commit safety scan: file size, anonymity (author / affiliation strings in tex/bib), hardcoded secrets, and invisible-Unicode carriers.

    146 GitHub stars~2.8k tokensUpdated 10 days ago
    Auto-check: notes

Questions about Wiki Grow

What does Wiki Grow do?

Auto-promote eligible project knowledge articles into Research Vault concepts with generated provenance metadata. Wiki Grow is an agent skill from flonat/flonat-research. Auto-promote eligible project knowledge articles into Research Vault concepts with generated provenance metadata.

When should I use Wiki Grow?

Wiki Grow fits situations like: running the scheduled knowledge-corpus promotion; previewing candidates interactively.

How do I install Wiki Grow in Claude Code?

Run `npx skills add flonat/flonat-research --skill wiki-grow -a claude-code`. Or copy the skill folder (skills/wiki-grow in flonat/flonat-research) into .claude/skills/wiki-grow in your project. Claude Code loads it when a task matches its description.

How do I install Wiki Grow in Codex?

Run `npx skills add flonat/flonat-research --skill wiki-grow -a codex`. Or copy the skill folder (skills/wiki-grow in flonat/flonat-research) into .agents/skills/wiki-grow in your project. Codex loads it when a task matches its description.

Can I use Wiki Grow in Cursor, Gemini CLI or GitHub Copilot?

Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add flonat/flonat-research --skill wiki-grow -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/wiki-grow, .gemini/skills/wiki-grow, .github/skills/wiki-grow and .opencode/skills/wiki-grow in your project.

What does Wiki Grow need to run?

Going by SKILL.md and its folder, Wiki Grow needs the command-line tools its instructions call (uv). Our summary lists: Python 3. Its frontmatter pre-approves these tools: Bash, Read.

Does Wiki Grow access the network?

SKILL.md contains no URLs. Its commands use uv, which can reach the network depending on how they are called. This is read from the text; nothing was executed.

Is Wiki Grow safe to install?

Our automated static check of SKILL.md found notes only (pre-approves every shell command (allowed-tools: bash)), nothing it rates as a warning. It is not a guarantee. Review the folder before installing.

What licence does Wiki Grow use?

Wiki Grow is published under the MIT licence (the repository's licence). It allows redistribution, so the full SKILL.md is shown on this page.

How many tokens does Wiki Grow use?

About 1.8k tokens (SKILL.md is roughly 7.2k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full.

What are the alternatives to Wiki Grow?

Skills that share tags, products or a category with Wiki Grow: JavaScript Concept Fact Checker (leonardomso/33-js-concepts, 67k stars), Beads Documentation Style Guide (gastownhall/beads, 28k stars), JavaScript Concept Page Workflow (leonardomso/33-js-concepts, 67k stars) and JS Concept Resource Curator (leonardomso/33-js-concepts, 67k stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.

Who maintains Wiki Grow?

flonat (a GitHub user) maintains it in flonat/flonat-research, which has 146 GitHub stars. The repository holds 83 skills in this directory. The repository was last updated on September 29, 2026.

Source: flonat/flonat-research on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.