A command for searching for recent news about topics in an LLM Wiki, a knowledge base about large language models.

It searches all topic groups, a named group, or keywords, and supports gap-filling searches for sparse groups, single-source topics, and outdated central pages.

It gathers new information for selected topics or topic groups and can identify gaps in the wiki's coverage.

How alfadur7/llm-wiki-newsroom configures Claude Code

Reuse

Borrowing it

Nothing to install: this file belongs to alfadur7/llm-wiki-newsroom. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.

Copy the file

curl -O https://raw.githubusercontent.com/alfadur7/llm-wiki-newsroom/main/.claude/commands/wiki-news.md

Clone the repo

git clone --depth 1 https://github.com/alfadur7/llm-wiki-newsroom

Made for: Claude Code.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

README.md

[![agentmods](https://agentmods.dev/badge/commands/alfadur7/llm-wiki-newsroom/wiki-news/github.svg)](https://agentmods.dev/commands/alfadur7/llm-wiki-newsroom/wiki-news)

Your own site

<a href="https://agentmods.dev/commands/alfadur7/llm-wiki-newsroom/wiki-news"><img src="https://agentmods.dev/badge/commands/alfadur7/llm-wiki-newsroom/wiki-news/github.svg" alt="Measured on agentmods" height="20"></a>

Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.

Your own site · 80×15

<a href="https://agentmods.dev/commands/alfadur7/llm-wiki-newsroom/wiki-news"><img src="https://agentmods.dev/badge/commands/alfadur7/llm-wiki-newsroom/wiki-news.svg" alt="Reviewed on agentmods" width="80" height="20"></a>

Per session 0 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.

When invoked 1,513 The whole file, excluding the scripts and references it only reads on demand.

Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.

Origin original No closer match found in the catalogue.

Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

Model	Per session	Once invoked
Fable 5.1	$0.00000	$0.01513
Opus 5	$0.00000	$0.00757
Sonnet 5	$0.00000	$0.00303
Haiku 4.5	$0.00000	$0.00151

Measured today against content hash bbf56a46a2cd, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-12, from the pricing page.

Security

Grade A, and why

wiki-news scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured today.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

.claude/commands/wiki-news.md · 81 lines

How it starts

The opening of the file, as written. The whole thing — 81 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Search for latest news related to the LLM Wiki's key topics.

Usage: /wiki-news [cluster|keyword | --gap [<slug>] [--no-filter]] — argument optional

If $ARGUMENTS is empty: show the usage below, then proceed with an all-cluster search (do not stop).

Usage: /wiki-news [cluster|keyword | --gap [<slug>] [--no-filter]]

Examples:
  /wiki-news                          # search all clusters (default)
  /wiki-news open-source-ai-definition  # a specific cluster (slug)
  /wiki-news open weights             # a specific cluster (name)
  /wiki-news DeepSeek open model      # free keyword
  /wiki-news --gap                    # gap fill: deterministic hub crawl → _inbox.md (Track A)
  /wiki-news --gap single-source      # enrich a single gap type
  /wiki-news --gap --no-filter        # disable the domain filter

The --gap slug is Track A only: single-source · stale-hub.

Traversal Pattern

Reading domain — the Reporter does an external search per cluster via parallel spawn, then (optionally) an ingest chain. It runs independent cluster queries concurrently in breadth-first parallel mode.

Phase	Cycle	Owner
Read wiki context	GROUND (own context)	Editor-in-Chief (read overview·index·`graph/_clusters.json`)
Build cluster queries	—	Editor-in-Chief
External WebSearch	GROUND breadth-first	Reporter multi-spawn (1 instance per cluster)
Duplicate check	—	Editor-in-Chief (`_source_map.json` matching)
Write report	—	Editor-in-Chief
(optional) inbox queue append	—	Editor-in-Chief (`_inbox.md`) — fetch·ingest is the separate `/wiki-ingest inbox`

Cluster slug list SoT: graph/_clusters.json::clusters[].slug (single SoT, varies with ingest accumulation).

Procedure (Editor-in-Chief Orchestration)

Read wiki context — wiki/overview.md (key themes·entities), wiki/index.md (cluster breakdown), graph/_clusters.json (clusters[].members·top_tags·source_assignments)
Build search queries:
- If a cluster argument is given — match slug/name against _clusters.json::clusters[], extract 3-5 member hubs with clear labels, combine the hub names (e.g. open-source-ai-definition → "OpenSourceInitiative Meta open weights AI definition 2026")
- If no argument — the top 5 clusters by _clusters.json size, combining 2-3 hubs each
- Add specific event·person keywords from the overview.md narrative (e.g. "DeepSeek open model release")
- Append the current year for recency
- Queries default to English; build Korean queries (for Korean news sources) only under WIKI_LANG=ko
Reporter parallel spawn — perform WebSearch per cluster (independent cluster queries run concurrently)
Duplicate check — match each result's title·URL against wiki/sources/_source_map.json (or grep wiki/sources/ for similar keywords)
Write report — markdown report of category·title·source·date·ingest status (new / already ingested)
Recommend top articles — suggest ingest + wait for the user's decision
After the user's decision — append the chosen article URLs to _inbox.md as # source=interactive (fetch·ingest is the separate /wiki-ingest inbox), or save the report to wiki/syntheses/

Read the full file on GitHub · 81 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

today Changed · -6 lines bbf56a46a2cd
12d ago First seen · 87 lines · 0 tokens per session scan A 3347bb168048

Subscribe to this mod's changes

wiki-news is a command published in the GitHub repository alfadur7/llm-wiki-newsroom (85 stars, last pushed yesterday), licensed MIT. It costs nothing until one of its globs matches a file; then it loads 1,513 tokens. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.

Claude Code Research agentic-ai ai-agents knowledge-base knowledge-graph local-first

llm-wiki-newsroom: Command for Claude Code

Borrowing it

Wrote this? Show the measurements

What it costs to keep this loaded

Grade A, and why

Nothing flagged

How it starts

Traversal Pattern

Procedure (Editor-in-Chief Orchestration)

What this file has done since we first saw it

Other commands, from other repositories

slm-loop

review

mt-onboard

add-source

choose-model

checklist