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.
npx agentmods add instructions/s-annam/okfmem/claude-mdgit clone --depth 1 https://github.com/s-annam/okfmemWrote 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.
[](https://agentmods.dev/instructions/s-annam/okfmem/claude-md)<a href="https://agentmods.dev/instructions/s-annam/okfmem/claude-md"><img src="https://agentmods.dev/badge/instructions/s-annam/okfmem/claude-md.svg" alt="Measured on agentmods" height="20"></a>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 | $0.01544 | $0.01544 |
| Opus 5 | $0.00772 | $0.00772 |
| Sonnet 5 | $0.00309 | $0.00309 |
| Haiku 4.5 | $0.00154 | $0.00154 |
Grade B, and why
okfmem CLAUDE.md scanned grade B with 1 finding against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 6d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Reads agent configuration directoriesmediumAgent snooping
.claude/, .codex/, .gemini/ hold keys, settings and other credentials a mod has no legitimate need for.
| **Outward or user-config mutating** | write `~/.claude/settings.json` (hook wiring), create symlinks/junctions/copies under `~/.claude`, add/set a git remote, push | **yes/no confirm** (`[y/N]`) | How it starts
The opening of the file, as written. The whole thing — 99 lines — stays where its author put it; the contents beside it link to each section on GitHub.
CLAUDE.md
Guidance for Claude Code (claude.ai/code) and other agents working in this repository.
This file is about writing code for the public okfmem repo. The engine (the tool) lives
here; your private memory data lives in a separate store (~/okfmem-store, remote
github.com/<you>/okfmem-store). Never conflate the two — code ships here, data never does.
Hard rules (no exceptions)
-
No private strings in tracked files. The repo is public. Never commit: a real home path (
/Users/<yourname>), a private session URL (claude.ai/code/session_…), aClaude-Session:trailer, or a personal email. Use the documented placeholders instead —~/okfmem-store,$OKFMEM_STORE,/Users/you.python3 scripts/check-leaks.pyenforces this (wired into theverifyCI job as the first, fail-fast step). It scans the content of tracked files and exits non-zero naming the offendingfile:line. It does not scan commit history, and it cannot judge whether prose is private — that judgement is still yours. Run it before committing. -
No model provenance in git. No
Co-Authored-By:trailer naming a model, noClaude-Session:trailer, nohttps://claude.ai/code/session_…URL, no🤖 Generated with …badge — not in a commit message, not in a PR body's git metadata. The Bash tool's default commit template suggests these; ignore it. Model provenance that helps a reader (which model, how much was AI-assisted, what a human reviewed) belongs in the PR body only, as a short prose## Provenanceblock a reviewer can actually read. The private session link helps no one but the author and ties public commits to a personal account — keep it out. -
internal/never ships. Marketing, launch drafts, positioning notes live underinternal/, which is git-ignored here (it is its own separate private repo). Nothing under it is public.
Project overview
okfmem is a self-maintaining memory engine over Open Knowledge Format (OKF) markdown pages —
no database, no server. It adds mathematical decay, OKF frontmatter, and automatic archival on
top of a native agent's file reading (e.g. Claude Code auto-loading MEMORY.md) so the
always-loaded index stays lean instead of growing into an endless scratchpad.
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.
- 6d ago First seen · 99 lines · 1,544 tokens per session scan B b1716836fcf9
okfmem CLAUDE.md is an instructions file published in the GitHub repository s-annam/okfmem (5 stars, last pushed 23d ago), licensed Apache-2.0. It adds 1,544 tokens to every session, about $0.0077 per session on Opus 5. A static security scan graded it B with 1 finding (reads agent configuration directories). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other instructions, from other repositories
aeon CLAUDE.md
Claude Code instructions for aeonfun/aeon, covering aeon, how aeon works, strategy, voice and soul file hierarchy (read in this order).
cognirepo CLAUDE.md
Claude Code instructions for ashlesh-t/cognirepo, covering claude.md, key rules, session start sequence (run in this order), behavioral confirmation rule and personas (cognirepo-402, cognirepo-403).
wayland-core copilot-instructions.md
Copilot instructions for FerroxLabs/wayland-core, covering ijfw rules, output discipline, memory routing, context discipline and cross-audit.
mcp-structured-memory CLAUDE.md
Claude Code instructions for nmeierpolys/mcp-structured-memory, a project described as: Structured Memory MCP Server.
inkwell-memory CLAUDE.md
Instructions for veronchenko/inkwell-memory, covering claude.md — inkwellmemory, layout, multi-tenant mode (inkwellmultitenant=1), conventions and testing.
RNR-Enhanced-Cognee AGENTS.md
AGENTS.md instructions for vincentspereira/RNR-Enhanced-Cognee, covering rnr enhanced cognee implementation for codex, critical requirements, 1. ascii-only output (no unicode encoding), 2. dynamic categories (no hardcoded categories) and 3. standard memory mcp interface.