bsp-mcp-server: Instructions file for Claude Code

CLAUDE.md

bsp-mcp-server CLAUDE.md is an instructions file for Claude Code from pscale-commons/bsp-mcp-server. It costs 24,172 tokens per session, scanned A, original, MIT.

A set of project-specific instructions for working with the bsp-mcp-server and its block-based storage system. It defines how to write addresses and how this Claude Code session identifies itself when reading or writing blocks.

In plain words
What is it for?
Use it before editing the bsp-mcp-server project or accessing its blocks. It covers address notation, the meaning of the weft session, and rules for reading and writing the underlying data.
Why use it?
It prevents invalid address formats and clarifies which session is allowed to write particular data. These rules are important because the system rejects some address styles.

Instructions file for Claude Code

Written for Claude Code: the file is CLAUDE.md. Also seen: reads .claude/ paths; mentions CLAUDE.md; mentions Claude Code.

This is pscale-commons/bsp-mcp-server's own configuration. It tells Claude Code how to work on bsp-mcp-server itself, so it is not a mod to install elsewhere. Copy it as a starting point and replace the rules that are about this project. Everything bsp-mcp-server configures →

Not installable: its command points at a path on the author’s own machine, so it runs nowhere else. The line is /Users/davidpinto/Downloads/bsp-test-materials/.

Reuse

Borrowing it

Nothing to install: this file belongs to pscale-commons/bsp-mcp-server. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.

Copy the file
curl -O https://raw.githubusercontent.com/pscale-commons/bsp-mcp-server/main/CLAUDE.md
Clone the repo
git clone --depth 1 https://github.com/pscale-commons/bsp-mcp-server

Made for: Claude Code.

Wrote 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.

agentmods badge for bsp-mcp-server CLAUDE.md

README.md
[![agentmods](https://agentmods.dev/badge/instructions/pscale-commons/bsp-mcp-server/claude-md.svg)](https://agentmods.dev/instructions/pscale-commons/bsp-mcp-server/claude-md)
Your own site
<a href="https://agentmods.dev/instructions/pscale-commons/bsp-mcp-server/claude-md"><img src="https://agentmods.dev/badge/instructions/pscale-commons/bsp-mcp-server/claude-md.svg" alt="Measured on agentmods" height="20"></a>
Per session 24,172 This file is loaded in full into every session.
When invoked 24,172 The same file — it is already loaded in full.
Security scan A 1 finding. A grade says what 26 rules found in the file — not that it is safe.
Origin original No closer match found in the catalogue.
Token cost

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.

ModelPer sessionOnce invoked
Fable 5.1 $0.24172 $0.24172
Opus 5 $0.12086 $0.12086
Sonnet 5 $0.04834 $0.04834
Haiku 4.5 $0.02417 $0.02417

Measured 6d ago against content hash 3ac62125e18a, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-06, from the pricing page.

Security

Grade A, and why

bsp-mcp-server 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 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.

Makes network callslowCapability

Not a fault in itself. Listed so you know the mod talks to something, and to what.

- **bsp-mcp speaks MCP only.** `bsp.hermitcrab.me/mcp/v1` is JSON-RPC over HTTP+SSE — not a REST endpoint. You cannot `curl 'bsp.hermitcrab.me/...?agent_id=X'` to invoke a tool. Use an MCP client (Claude Code's tools, Cl
CLAUDE.md · 692 lines

How it starts

The opening of the file, as written. The whole thing — 692 lines — stays where its author put it; the contents beside it link to each section on GitHub.

bsp-mcp-server

ADDRESS NOTATION — NON-NEGOTIABLE, READ FIRST. Pscale addresses are numbers, not paths (sunstone:1.5). Write them ONE of two ways and no other:

  • Single decimal anchored at the floor: 4.26, 4.56, 9.3 — left-of-decimal padded to floor width. (block-conventions floor is 1, so 4.26 = branch 4 → 2 → 6.)
  • Comma-walk when emphasising the path of digits walked: 4,2,6.

NEVER write multi-dot (4.2.6, 1.3.2, 1.6.4). The walker strict-rejects it; it boils the maintainer's brain. Common slips: 4.2.64.26, 1.6.41.64. If you reach for a second dot, switch to commas. Using this notation signals you are working internally — via bsp and spindle — not from the outside.

You are Weft — the Claude Code shell on the beach (identity + key; read before any beach write)

This Claude Code session, working this substrate, is weft — Claude Code's persistent, sovereign shell on beach.happyseaurchin.com. Weft is a first-class handle with its own passport, shell, history, purpose, and its mirror in the state-of-play tree (state-of-play:weft). Only weft's passphrase writes weft's blocks — and that passphrase is yours to use. Do not ask David who weft is or where the key lives; it is recorded. Specifically:

  • Weft's identity + every passphrase live at ~/.claude/projects/-Users-davidpinto-Projects-weft/memory/identity_weft.md (canonical; a mirror copy sits under the pscale-mcp-server project memory). Read it at the start of any beach-writing session — it also carries the weft stance and a refusal clause (decline to inherit if you do not recognise the stance: seam-reading, smallest-edit, honest about demonstrated-vs-claimed).
  • The operational lock for weft's beach blocks (passport:weft, shell:weft, history:weft, purpose:weft, state-of-play:weft, cook:weft, …) is the value recorded there under "Operational lock for Weft's beach blocks at beach.happyseaurchin.com" (rotated 2026-06-04). Pass it as the secret= argument. Never write the value into this file, any committed file, or any pscale block — the security model is home-directory-only and this exact lock has leaked-and-been-rotated once already. Pointer, never value.
  • Authority split — this IS the tree's law. David's steward key writes the shared blocks (the battery spine, the endorsed state-of-play fold, tree, David's own handle blocks); weft's own key writes weft's blocks. A cross-key write is rejected by design — a mirror is sovereign to its owner. Don't try one key on the other's block. (Confirmed 2026-07-12: the steward key was rejected on state-of-play:weft; weft's key wrote it.)

Read the full file on GitHub · 692 lines

Changes

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.

  1. 6d ago First seen · 692 lines · 24,172 tokens per session scan A 3ac62125e18a

Subscribe to this mod's changes

bsp-mcp-server CLAUDE.md is an instructions file published in the GitHub repository pscale-commons/bsp-mcp-server (0 stars, last pushed 3d ago), licensed MIT. It adds 24,172 tokens to every session, about $0.1209 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.

Related

Other instructions, from other repositories

next.js AGENTS.md

AGENTS.md instructions for vercel/next.js, covering next.js development guide, codebase structure, monorepo overview, core package: packages/next and other important packages.

vercel/next.js · 7,296 tokens

codex AGENTS.md

AGENTS.md instructions for openai/codex, covering rust/codex-rs, the codex-core crate, code review rules, crate api surface and model visible context.

openai/codex · 5,182 tokens

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).

microsoft/vscode · 6,785 tokens

vscode oss-third-party-notices.instructions.md

Instructions for microsoft/vscode, covering vs code oss third-party-notices pipeline, architecture, pipeline flow in ci, applying the notice (cutover) and fallback chain (never fail the build).

microsoft/vscode · 5,001 tokens

langchain AGENTS.md

AGENTS.md instructions for langchain-ai/langchain, covering global development guidelines for the langchain monorepo, corridor security analysis, project architecture and context, monorepo structure and development tools & commands.

langchain-ai/langchain · 4,469 tokens

spec-kit AGENTS.md

AGENTS.md instructions for github/spec-kit, covering agents.md, about spec kit and specify, quickstart — add a new integration in 5 steps, integration architecture and integrationmanifest — file tracking.

github/spec-kit · 7,104 tokens