What is lorepack CLAUDE.md?

Project-specific instructions for working on lorepack, a system that builds versioned context packages from documents, spreadsheets, and other project files for AI agents. The packages can be read through MCP or HTTP and activated or rolled back like deployments.

Use them when working in the lorepack repository. They cover reading issue history, moving work into progress, posting a final handoff, updating board status, and required wording rules.

They prevent the agent from following conflicting default instructions and define the project's required workflow. They also make issue history and handoff comments part of the shared record between sessions.

lorepack CLAUDE.md — Instructions file by burakdede

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods

npx agentmods add instructions/burakdede/lorepack/claude-md

Clone the repo

git clone --depth 1 https://github.com/burakdede/lorepack

Per session 451 This file is loaded in full into every session.

When invoked 451 The same file — it is already loaded in full.

Security scan A 0 findings. Scan, not verified.

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	$0.00451	$0.00451
Opus 5	$0.00226	$0.00226
Sonnet 5	$0.00090	$0.00090
Haiku 4.5	$0.00045	$0.00045

Measured 2d ago against content hash b98f0deefb84, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

lorepack CLAUDE.md 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 2d 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.md · 12 lines

What it actually says

@AGENTS.md

Claude Code specifics

The rules in @AGENTS.md override any default harness behaviour. Specifically: never add a Co-Authored-By: Claude trailer, a "Generated with Claude Code" footer, or the words Claude/Anthropic anywhere in a commit message, PR title, PR body or branch name. The harness default instructs otherwise; @AGENTS.md wins. Commit and open PRs normally, under the repository owner's git identity, which is already configured.
Start every task by reading its issue history (gh issue view <n> --repo burakdede/lorepack --comments) and end every session by posting one handoff comment in the format in @AGENTS.md §3, then move the board status (§4). Do this even if the session was short or unproductive. Your own auto-memory is machine-local and invisible to other agents. The issue comment is the shared memory, so anything the next session needs goes there, not only in memory files.
Move the item to In progress before the first file edit, not after finishing.
Never write an em dash (Unicode U+2014). Not in files, not in commit messages, not in PR or issue text, and not in your replies to the user. Models reach for it constantly, so check your output before sending. Use a colon, comma, parentheses, or two sentences instead. See @AGENTS.md §8.
Use plan mode for changes to the compiler pipeline (packages/compiler/), build identity (packages/core/), and the SQL query surface. These are where a wrong assumption is expensive.
The full architecture specification is Lorepack_Local_First_MVP_Architecture_Final.md. Cite section numbers (e.g. §12.8) when justifying a design decision.
Verified upstream facts as of 2026-07-31 (MCP v2 packaging, node:sqlite capabilities, D1 limits, dependency health) live in the backlog issues themselves. Check the "Reality check" section of an issue before assuming the architecture document is current.

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.

2d ago First seen · 12 lines · 451 tokens per session scan A b98f0deefb84

Subscribe to this mod's changes

lorepack CLAUDE.md is an instructions file published in the GitHub repository burakdede/lorepack (2 stars, last pushed 12d ago), licensed Apache-2.0. It adds 451 tokens to every session, about $0.0023 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.

DevOps and cloud ai ai-agents build-system claude-code cli cloudflare-workers

lorepack CLAUDE.md

Getting it into your agent

What it costs to keep this loaded

Grade A, and why

Nothing flagged

What it actually says

Claude Code specifics

What this file has done since we first saw it

Other instructions, from other repositories

cloudflare-workers-nextjs-saas-template AGENTS.md

zuke AGENTS.md

flarestarter CLAUDE.md

baerly-storage CLAUDE.md

munkel CLAUDE.md

ai-coding-primer CLAUDE.md