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/llm-coding/semantic-anchors/claude-mdgit clone --depth 1 https://github.com/LLM-Coding/Semantic-AnchorsWhat 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 | $0.09005 | $0.09005 |
| Opus 5 | $0.04503 | $0.04503 |
| Sonnet 5 | $0.01801 | $0.01801 |
| Haiku 4.5 | $0.00901 | $0.00901 |
Grade A, and why
Semantic-Anchors 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 3d 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.
How it starts
The opening of the file, as written. The whole thing — 725 lines — stays where its author put it; the contents beside it link to each section on GitHub.
CLAUDE.md
This file provides guidance to Claude Code (claude.ai/code) when working with code in this repository.
Repository Purpose
This repository is a curated catalog of semantic anchors - well-defined terms, methodologies, and frameworks that serve as reference points when communicating with Large Language Models (LLMs). Semantic anchors act as shared vocabulary that triggers specific, contextually rich knowledge domains within an LLM's training data.
Current Status: The repository is undergoing a major redesign to become an interactive, bilingual website with treemap visualization, role-based filtering, and automated contribution workflow.
First-Time Setup
After cloning, activate the shared Git hooks with one command (works on Unix/macOS and Windows Git Bash):
git config core.hooksPath .githooks
This installs the pre-commit hook that automatically syncs skill/ into plugins/semantic-anchors/skills/ on every commit. No other tools are required.
Git Workflow & Fork Setup
Development takes place in the fork until PRD implementation is complete.
Git Remotes:
origin→raifdmueller/Semantic-Anchors(fork for development)upstream→LLM-Coding/Semantic-Anchors(original repository)
Workflow:
- All feature branches are created in the fork (
origin) - PRs are created within the fork for review
- After PRD implementation is complete, changes will be merged back to upstream
- Use
git push origin <branch>for all development work - Use
git fetch upstreamto sync with original repository if needed
Issue References:
- Issues are tracked in upstream repository: https://github.com/LLM-Coding/Semantic-Anchors/issues
- Reference issues in commits:
feat: implement X (#42)
Git Worktrees for Parallel Team Development
When using agent teams, use Git Worktrees to enable true parallel development:
Setup:
# Create worktree directory
mkdir -p ~/projects/Semantic-Anchors-worktrees
# Create worktree for each feature branch
git worktree add ../Semantic-Anchors-worktrees/feature-<name> -b feature/<name>
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.
- 3d ago First seen · 725 lines · 9,005 tokens per session scan A 63361d4e3a59
Semantic-Anchors CLAUDE.md is an instructions file published in the GitHub repository LLM-Coding/Semantic-Anchors (463 stars, last pushed 2d ago), licensed Apache-2.0. It adds 9,005 tokens to every session, about $0.0450 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
ai-dossier CLAUDE.md
Claude Code instructions for imboard-ai/ai-dossier, a project described as: Universal standard for LLM-executable automation.
my-skills CLAUDE.md
Claude Code instructions for mizzy/my-skills, covering claude.md, what this repo is, bump the plugin version on every skill change, the skills form one workflow chain and codex is the execution model.
memvid-rag CLAUDE.md
Claude Code instructions for fa-ina-tic/memvid-rag, covering claude code context, project structure, technology stack, key commands and development notes.
resume-buddy CLAUDE.md
Claude Code instructions for SiamRahman29/resume-buddy, covering resume buddy, the model, workflow rules for the agent, skills and browser (vendor/browse, used by resume-scrape).
repolore AGENTS.md
AGENTS.md instructions for emyann/repolore, covering repolore — agent notes and project wiki (llm-maintained).
instancez CLAUDE.md
Claude Code instructions for instancez/instancez, a project described as: An LLM-friendly, single-binary, drop-in replacement to Supabase.