agentmods.dev

agent-logbook: Instructions file for Claude Code

CLAUDE.md

agent-logbook CLAUDE.md is an instructions file for Claude Code from Rajwantmishra/agent-logbook. It costs 464 tokens per session, scanned C, original, MIT.

Project instructions for using a memory MCP server as long-term project memory. The agent’s context window is temporary working space, while the memory server stores information for later tasks.

In plain words
What is it for?
Use them to guide an agent through recalling relevant history at the start of work and saving durable results afterward. They specify memory entry types, sentence limits, stable subject names, and conflict handling.
Why use it?
They prevent important decisions, facts, follow-up tasks, and questions from being lost between sessions. They also define when to read from and write to the memory store.

Instructions file for Claude Code

Written for Claude Code: the file is CLAUDE.md. Also seen: mentions CLAUDE.md.

This is Rajwantmishra/agent-logbook's own configuration. It tells Claude Code how to work on agent-logbook 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 agent-logbook configures →

not rated 0repo 3mo ago C scan 464 tokens original MIT
View source ↗ Rajwantmishra/agent-logbook
Reuse

Borrowing it

Nothing to install: this file belongs to Rajwantmishra/agent-logbook. 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/Rajwantmishra/agent-logbook/master/CLAUDE.md
Clone the repo
git clone --depth 1 https://github.com/Rajwantmishra/agent-logbook

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 agent-logbook CLAUDE.md

README.md
[![agentmods](https://agentmods.dev/badge/instructions/rajwantmishra/agent-logbook/claude-md/github.svg)](https://agentmods.dev/instructions/rajwantmishra/agent-logbook/claude-md)
Your own site
<a href="https://agentmods.dev/instructions/rajwantmishra/agent-logbook/claude-md"><img src="https://agentmods.dev/badge/instructions/rajwantmishra/agent-logbook/claude-md/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 agent-logbook CLAUDE.md

Your own site · 80×15
<a href="https://agentmods.dev/instructions/rajwantmishra/agent-logbook/claude-md"><img src="https://agentmods.dev/badge/instructions/rajwantmishra/agent-logbook/claude-md.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 464 This file is loaded in full into every session.
When invoked 464 The same file — it is already loaded in full.
Security scan C 1 finding. 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 $0.00464 $0.00464
Opus 5.5 $0.00186 $0.00186
Sonnet 5.5 $0.00093 $0.00093
Haiku 4.5 $0.00046 $0.00046

Measured 2mo ago against content hash 1508ce8abfaa, method: parsed. Prices are Anthropic first-party input rates as of 2026-10-07, from the pricing page.

Security

Grade C, and why

agent-logbook CLAUDE.md scanned grade C with 1 finding 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 2mo 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.

Hidden instructionshighPrompt injection

Directives inside HTML comments, invisible characters or bidirectional overrides are read by the model and not by the person reviewing the file.

<!-- Paste this section into your project's CLAUDE.md or system instructions. -->
CLAUDE.md · 22 lines

What it actually says

Memory protocol

This project has a long-term memory served by the memory MCP server. The context window is working memory only; durable knowledge lives in the memory store. Follow this protocol:

  1. At the start of every task, call recall(query) with a few keywords describing the task, before doing any work. If open_items_count > 0 and you are planning, also call list_open().

  2. After completing any work block that produced a decision, a durable fact, a new task, or an open question, call remember(type, content, entity). Rules for what you store:

    • content is ONE or TWO self-contained sentences, declarative, understandable without any conversation context. Distill — never paste transcript or code blocks.
    • entity is a short stable slug for the subject (e.g. auth, db-choice, deploy-pipeline). Reuse existing entity slugs when the subject is the same.
    • Types: decision (choices made and why), fact (durable truths about the project), task (work to do later), question (unresolved), note (everything else). Prefer decision/fact over note.

  3. If remember returns a conflict, never force silently. If the new information clearly replaces the old, call supersede(old_id, new_content). If it is genuinely ambiguous, ask the user which stands. Use force=true only when both memories are simultaneously true.

  4. When the user says "actually..." or changes a prior decision, find the old memory (recall it if needed) and supersede it — do not create a parallel memory.

  5. When a task or question is resolved, call set_status(id, "done").

  6. Do not store: secrets, API keys, credentials, personal data, transient debugging state, or anything the user asks to keep out of memory.

  7. Memory is the source of truth for past decisions. If your recollection from context conflicts with a recalled memory, trust the memory or ask.

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. 2mo ago First seen · 22 lines · 464 tokens per session scan C 1508ce8abfaa

Subscribe to this mod's changes

agent-logbook CLAUDE.md is an instructions file published in the GitHub repository Rajwantmishra/agent-logbook (0 stars, last pushed 3mo ago), licensed MIT. It adds 464 tokens to every session, about $0.0019 per session on Opus 5.5. A static security scan graded it C with 1 finding (hidden instructions). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.

Memory and context
Related

agent-logbook CLAUDE.md compared with instructions from other repositories

agent-logbook CLAUDE.md next to 6 instructions from other repositories
ModTypeStarsLast pushLicenceScanTokensWorks with
agent-logbook CLAUDE.mdRajwantmishra/agent-logbookInstructions file03mo agoMITC464Claude Code
vscode buildNext.instructions.mdmicrosoft/vscodeInstructions file195kyesterdayMITA9,719GitHub Copilot
next.js AGENTS.mdvercel/next.jsInstructions file143kyesterdayMITA7,461Codex, OpenCode
langchain AGENTS.mdlangchain-ai/langchainInstructions file149kyesterdayMITA4,680Codex, OpenCode
vscode oss-third-party-notices.instructions.mdmicrosoft/vscodeInstructions file195kyesterdayMITA5,516GitHub Copilot
vscode copilot-instructions.mdmicrosoft/vscodeInstructions file195kyesterdayMITA3,093GitHub Copilot
deepseek-harness AGENTS.mddeepseek-ai/deepseek-harnessInstructions file274k2d agoMITA4,313Codex, OpenCode