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/harmonicabot/harmonica-chat/claude-mdgit clone --depth 1 https://github.com/harmonicabot/harmonica-chatWhat 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 | $0.01443 | $0.01443 |
| Opus 5 | $0.00722 | $0.00722 |
| Sonnet 5 | $0.00289 | $0.00289 |
| Haiku 4.5 | $0.00144 | $0.00144 |
Grade A, and why
harmonica-chat CLAUDE.md scanned grade A 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 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.
Makes network callslowCapability
Not a fault in itself. Listed so you know the mod talks to something, and to what.
**Version auto-update.** Line 6 of SKILL.md is `<!-- harmonica-chat vX.Y.Z -->`. On each run, the skill fetches the latest from GitHub and notifies users if outdated. Bump this version comment when making changes. (Pre-v How it starts
The opening of the file, as written. The whole thing — 69 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.
Project Overview
harmonica-chat is a Claude Code skill (/harmonica-chat) that designs, creates, and manages Harmonica deliberation sessions. It's distributed as a thin entry router (SKILL.md) plus 11 subcommand reference files (reference/*.md) that users install to ~/.claude/skills/harmonica-chat/.
The skill requires the harmonica-mcp server for API access.
Repository Structure
SKILL.md— Entry router (~120 lines). Frontmatter, prerequisites, session design laws, argument routing, commands menu. This file MUST stay thin; subcommand logic lives inreference/.reference/— 11 subcommand reference files. Loaded on demand by the entry router.design.md— Mode 1: guided session design, 14 stepsaccelerated.md— Mode 2: topic-upfront flow + project-aware creationstatus.md,check.md,summary.md,edit.md,review.md,follow-up.md— lifecycle commandsinvitation.md— post-creation invitation flow (join URL, draft message, community feed)expertise.md— soft-nudge session design heuristics (goal quality, context calibration, cross-pollination, critical questions)
Templates are NOT in reference/ — they're fetched at runtime via mcp__harmonica__list_templates (harmonica-mcp ≥ 0.11.0). The Harmonica admin panel is the single source of truth; the skill never hardcodes a template list. See HAR-1227.
install.sh/install.ps1— One-line installers. Download SKILL.md + all reference files. v3.0.0+ also migrates the legacy v2.x slash-command install (removes~/.claude/commands/harmonica-chat.mdif present).docs/plans/— Design docs for major rewrites (reference only, not part of the product).
How the Skill Works
Triggers when the user invokes /harmonica-chat or asks Claude Code to do something matching the description ("design a Harmonica session", "check on the brainstorming session", etc.).
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 · 69 lines · 1,443 tokens per session scan A 3a7106c1df86
harmonica-chat CLAUDE.md is an instructions file published in the GitHub repository harmonicabot/harmonica-chat (2 stars, last pushed 15d ago), licensed MIT. It adds 1,443 tokens to every session, about $0.0072 per session on Opus 5. A static security scan graded it A with 1 finding (makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other instructions, from other repositories
Claude-Code-Projects-Index CLAUDE.md
Claude Code instructions for danielrosehill/Claude-Code-Projects-Index, covering claude code repos index - repository instructions, purpose, repository structure, astro site architecture and modular category system.
ah AGENTS.md
Instructions for nihen/ah, covering ah - agent history, build & install, architecture, key design decisions and configuration.
catalyst AGENTS.md
AGENTS.md instructions for coalesce-labs/catalyst, covering agents.md, what this repository is, agent philosophy, build & test and key principles.
notebooklm-claude-workflows CLAUDE.md
Instructions for CreatmanCEO/notebooklm-claude-workflows, covering notebooklm-claude-workflows — claude.md (level 1), status: active, architecture, key files (when touching) and critical rules — when editing this repo.
apex-spec-system CLAUDE.md
Claude Code instructions for moshehbenavraham/apex-spec-system, covering apex spec system - claude code plugin skill, when updating version, update it in all the following files, apex infinite cli and details - see @skills/spec-workflow/skill.md.
shipit AGENTS.md
Instructions for nikzlabs/shipit, a project described as: Self-hosted, chat-driven IDE for running coding agents through your real engineering loop — Claude Code or Codex, parallel sandboxed containers, live preview, and the GitHub loop inline.