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/mnemehq/mneme/claude-mdgit clone --depth 1 https://github.com/MnemeHQ/mnemeWrote 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/mnemehq/mneme/claude-md)<a href="https://agentmods.dev/instructions/mnemehq/mneme/claude-md"><img src="https://agentmods.dev/badge/instructions/mnemehq/mneme/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 | $0.00929 | $0.00929 |
| Opus 5 | $0.00464 | $0.00464 |
| Sonnet 5 | $0.00186 | $0.00186 |
| Haiku 4.5 | $0.00093 | $0.00093 |
Grade A, and why
mneme 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 yesterday.
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.
How it starts
The opening of the file, as written. The whole thing — 76 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Mneme Repo Instructions
Branch Naming
Use conventional prefixes: feat/, fix/, site/, ci/, docs/, refactor/. Keep slugs short, kebab-case, no random suffixes unless required for uniqueness.
Known exception — auto-generated worktree branches: Claude Code's --worktree flag emits claude/<adjective>-<noun>-<hash> branch names. This is a hard-coded harness behavior with no configuration override. These branches are acceptable during development. Before opening a PR, rename to follow the taxonomy where practical. The squash commit title on main must follow the taxonomy regardless of source branch name. See ADR-010.
Agent Execution Context Guard
Worktrees in this repo share one .git. An agent that starts work in the wrong checkout inherits that checkout's branch and commits to it. Branch selection is therefore not agent discretion: the orchestrating task declares the expected context; the agent only verifies it.
At the start of every agent task:
- Run
git rev-parse --show-toplevelandgit branch --show-current. - Verify the repository root and branch match the task's declared worktree root and task branch.
Immediately before every git commit:
- Repeat the same assertion.
- Abort on any mismatch or on a detached HEAD.
The repo-owned checker automates this:
python scripts/check_worktree_context.py --expected-root <worktree-root> --expected-branch <task-branch>
It is read-only and fails closed (exit 1) with expected-vs-actual output on any mismatch. Never pass it values you did not receive from the task definition.
Provisioning and the automatic commit gate:
- Create each task worktree with
python scripts/new_task_worktree.py <branch>— it creates the branch from currentorigin/mainand writes.mneme/task_context.jsoninside the worktree. - When run without explicit arguments, the checker reads
.mneme/task_context.json; the versioned pre-commit hook (scripts/githooks/pre-commit) runs it before every commit in worktrees that have a context file. - One-time setup for the hook:
git config core.hooksPath scripts/githooks(installs bothpre-commitandpre-push).
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.
- yesterday Changed · +2 lines · +57 tokens per session 94dcde192f71
- 5d ago First seen · 74 lines · 872 tokens per session scan A 1579ced3da66
mneme CLAUDE.md is an instructions file published in the GitHub repository MnemeHQ/mneme (20 stars, last pushed today), licensed MIT. It adds 929 tokens to every session, about $0.0046 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 2026-08-30.
Other instructions, from other repositories
snip CLAUDE.md
Claude Code instructions for edouard-claude/snip, covering claude.md, project overview, key concept, repository structure and architecture.
opensrc AGENTS.md
Instructions for vercel-labs/opensrc, covering agents.md, monorepo structure, cli (packages/opensrc), docs (apps/docs) and releasing.
lid AGENTS.md
Instructions for jszmajda/lid, covering agents.md, repository purpose, structure, plugin architecture (claude code) and other agentic coding tools.
detrix CLAUDE.md
Instructions for flashus/detrix, covering claude.md, rust development, workflows, code audit workflow and task completion.
nanocoder CLAUDE.md
Claude Code instructions for Nano-Collective/nanocoder, covering claude.md, development commands, build and run, testing (run before committing) and individual test commands.
nanocoder AGENTS.md
AGENTS.md instructions for Nano-Collective/nanocoder, covering agents.md, project overview, architecture, key files and development commands.