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/kimsanguine/hplan/claude-mdgit clone --depth 1 https://github.com/kimsanguine/hplanWrote 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/kimsanguine/hplan/claude-md)<a href="https://agentmods.dev/instructions/kimsanguine/hplan/claude-md"><img src="https://agentmods.dev/badge/instructions/kimsanguine/hplan/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 | $0.01379 | $0.01379 |
| Opus 5 | $0.00690 | $0.00690 |
| Sonnet 5 | $0.00276 | $0.00276 |
| Haiku 4.5 | $0.00138 | $0.00138 |
Grade A, and why
hplan 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 5d 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 — 125 lines — stays where its author put it; the contents beside it link to each section on GitHub.
CLAUDE.md — hplan Project Rules (ADK L1: Memory)
ADK Layer 1 — Memory: This file is loaded into every Claude Code session automatically. It encodes the 9 behavioral disciplines that prevent the most common AI-assisted PM failures. Clone this repo and these rules apply immediately — no manual setup required.
Customize the hplan Context section below for your project. Keep the 9 Rules as-is.
hplan Context
This project uses hplan — the Product Build Gate for AI Agents.
| Layer | What | File |
|---|---|---|
| L1 Memory | Project rules (this file) | CLAUDE.md |
| L2 Skills | Auto-invoked PM disciplines | hplan/, discover/, architect/, deliver/, operate/ |
| L3 Hooks | Gate enforcement at tool time | hooks/ → .claude/settings.json |
| L4 Subagents | Task-sequential subagent dispatch + gates | deliver/skills/conductor/ |
| L5 Plugins | Marketplace distribution | hplan/PLUGIN.md |
Active harness directory: harness/
Gate status: Run /harness-doctor to verify hook installation.
9 Behavioral Rules
Rule 1 — Think Before Coding
State assumptions explicitly in words. Stop and ask rather than guessing if ambiguous. If there are two or more interpretations, show both and let the user choose. Before adding new keys/fields to external systems (settings.json · API · DB), verify the official schema first — never guess key names.
Check: Are assumptions/interpretations stated at the top of the response? Did I proceed despite remaining ambiguity? If yes → violation.
Rule 2 — Simplicity First
Write the minimum code that solves the request. No speculative features. No abstractions for one-off code. No try/except for scenarios that can't happen.
Check: Did I add branches that won't be called? Did I build the next task in advance? If yes → violation.
Rule 3 — Surgical Changes
Touch only what's needed. No "improving" adjacent code, comments, or formatting. No refactoring things that aren't broken. Preserve existing style.
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.
- 5d ago First seen · 125 lines · 1,379 tokens per session scan A 99f13eb539c2
hplan CLAUDE.md is an instructions file published in the GitHub repository kimsanguine/hplan (2 stars, last pushed 20d ago), licensed MIT. It adds 1,379 tokens to every session, about $0.0069 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 2026-08-31.
Other instructions, from other repositories
product-pipeline-public CLAUDE.md
Instructions for lenar-amirov/product-pipeline-public, covering product discovery — pm copilot, 🛑 stop — read this before responding to any user message, tracker integration, quick setup and first launch — value first, setup later.
unica CLAUDE.md
Claude Code instructions for IngvarConsulting/unica: See AGENTS.md for entry points, source-of-truth ordering, search hygiene, and development rules. That file applies to Claude Code unchanged.
erpaval CLAUDE.md
Instructions for theagenticguy/erpaval, covering erpaval-plugin, repo structure, version management, tools and hooks.
ed3d-plugins CLAUDE.md
Claude Code instructions for ed3dai/ed3d-plugins, covering ed3d-plugins, conventions, task invocations use xml syntax and version updates require marketplace and changelog sync.
openehr-assistant-plugin copilot-instructions.md
Instructions for Cadasto/openehr-assistant-plugin: Follow the repository guidelines in AGENTS.md for coding standards, project structure, MCP conventions, and developer workflows.
vscode buildNext.instructions.md
Working notes and architecture documentation for the new esbuild-based build system in build/next. Use when making changes to the new build pipeline (transpile/bundle commands, NLS plugin, source-map handling, resource copying, or self-hosting watch tasks).