obsidian-llm-wiki: Command for Claude Code

.claude/commands/pre-review.md

pre-review is a command for Claude Code from mickeytony0215-png/obsidian-llm-wiki. It costs 45 tokens per session (1,383 once invoked), scanned A, original, MIT.

A batch step that prepares paper notes for human review by comparing their claims with the source PDFs. It saves a preliminary report in each note but does not mark the note as reviewed.

In plain words
What is it for?
Use it to pre-audit selected unreviewed paper notes, check claims against PDF evidence, flag logical or data problems, and prepare reports for the later review command. The notes remain unreviewed until a person confirms them.
Why use it?
It handles the repetitive comparison work before a person performs the final review. The reviewer can then confirm an item-by-item checklist instead of starting with an empty review.

Command for Claude Code

Written for Claude Code: installed under .claude/.

This is mickeytony0215-png/obsidian-llm-wiki's own configuration. It tells Claude Code how to work on obsidian-llm-wiki itself, so it is not a mod to install elsewhere. Copy it as a starting point and replace the rules that are about this project. Everything obsidian-llm-wiki configures →

Reuse

Borrowing it

Nothing to install: this file belongs to mickeytony0215-png/obsidian-llm-wiki. 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/mickeytony0215-png/obsidian-llm-wiki/main/.claude/commands/pre-review.md
Clone the repo
git clone --depth 1 https://github.com/mickeytony0215-png/obsidian-llm-wiki

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.

agentmods badge for pre-review

README.md
[![agentmods](https://agentmods.dev/badge/commands/mickeytony0215-png/obsidian-llm-wiki/pre-review/github.svg)](https://agentmods.dev/commands/mickeytony0215-png/obsidian-llm-wiki/pre-review)
Your own site
<a href="https://agentmods.dev/commands/mickeytony0215-png/obsidian-llm-wiki/pre-review"><img src="https://agentmods.dev/badge/commands/mickeytony0215-png/obsidian-llm-wiki/pre-review/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.

agentmods 80×15 button for pre-review

Your own site · 80×15
<a href="https://agentmods.dev/commands/mickeytony0215-png/obsidian-llm-wiki/pre-review"><img src="https://agentmods.dev/badge/commands/mickeytony0215-png/obsidian-llm-wiki/pre-review.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 45 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 1,383 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.

ModelPer sessionOnce invoked
Fable 5.1 $0.00045 $0.01383
Opus 5 $0.00023 $0.00691
Sonnet 5 $0.00009 $0.00277
Haiku 4.5 $0.00005 $0.00138

Measured 12d ago against content hash 4fa58e9d3143, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-12, from the pricing page.

Security

Grade A, and why

pre-review 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 12d ago.

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/pre-review.md · 93 lines

How it starts

The opening of the file, as written. The whole thing — 93 lines — stays where its author put it; the contents beside it link to each section on GitHub.

/pre-review — batch pre-audit (a throughput part for the review gate)

Runs the automatable part of /review-note (load evidence, decompose claims, claim-by-claim PDF check, logical-consistency check, data-credibility red flags) ahead of time, in batch, and caches the result inside each note. When the user later runs /review-note, it starts directly from a confirmation checklist instead of from scratch — cutting the per-paper human review time from roughly 30–40 minutes down to roughly 10.

Hard rules (gate semantics never change)

  • Never modify review_status / reviewed_date / reviewed_by / review_notes. After /pre-review runs, the note is still unreviewed and /ingest-to-wiki still hard-blocks it.
  • The pre-review report is draft evidence, not a review verdict. Only the user, confirming item-by-item in /review-note, can change review_status.
  • Do not touch any existing content in the note body — only add or update the ## Pre-review report section.

Triggers

  • /pre-review — auto-selects the top 3 notes (by mtime, newest first) from the core tier of the triage queue (see /read-paper's triage rules and wiki/pending-review.md)
  • /pre-review <N> — take the top N from the core tier
  • /pre-review <note> [<note>...] — target specific notes (any triage tier)

Procedure (process one note fully before starting the next)

Step 1: Load PDF evidence (same as /review-note step 1)

  1. Read the note and locate its paired PDF (via frontmatter or a filename match under raw/papers/).
  2. Load full-text evidence. If a chunk-level retrieval backend (see /read-paper step 0 / /paper-ask) is available, prefer it — its evidence is page-anchored and its tables are pre-structured, which is more reliable than re-reading the PDF by eye.
    • Backend not available and the user doesn't want to bring it up: skip this note and report it (the value of pre-review is chunk-level comparison; a degraded pre-review defeats the purpose) — do not silently fall back to a raw PDF read.
  3. PDF not found: skip the note and report it under "could not pre-review."

Read the full file on GitHub · 93 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.

  1. 12d ago First seen · 93 lines · 45 tokens per session scan A 4fa58e9d3143

Subscribe to this mod's changes

pre-review is a command published in the GitHub repository mickeytony0215-png/obsidian-llm-wiki (2 stars, last pushed 2mo ago), licensed MIT. It adds 45 tokens to every session and 1,383 once invoked, about $0.0002 per session on Opus 5. 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 2026-08-31.