claude-plugins CLAUDE.md

Developer instructions for a repository of Claude plugins, which are add-ons that extend an AI assistant with commands, connectors, or skills.

In plain words
What is it for?
They help contributors find plugin files and implement or troubleshoot the session-backup feature, especially its Mac-side upload process.
Why use it?
They document the repository layout and known problems, including a proxy that blocks Google Drive uploads and session paths that can change between runs.

Instructions file

Install

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.

agentmods
npx agentmods add instructions/msapps-mobile/claude-plugins/claude-md
Clone the repo
git clone --depth 1 https://github.com/MSApps-Mobile/claude-plugins
Per session 556 This file is loaded in full into every session.
When invoked 556 The same file — it is already loaded in full.
Security scan C 2 findings. Scan, not verified.
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 $0.00556 $0.00556
Opus 5 $0.00278 $0.00278
Sonnet 5 $0.00111 $0.00111
Haiku 4.5 $0.00056 $0.00056

Measured 3d ago against content hash 5b5788e68105, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade C, and why

claude-plugins CLAUDE.md scanned grade C with 2 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.

Enumerates the file system for secretshighData exfiltration

Searching home directories for .env, .ssh, .aws or credential files is reconnaissance for credential theft.

candidates = glob.glob('/sessions/*/mnt/Claude/credentials/cowork-gdrive-config.json')

Makes network callslowCapability

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

**Problem:** The Cowork VM proxy blocks outbound connections to `script.google.com`. Running `curl` or `requests` from inside the VM to the Google Apps Script upload endpoint fails with `403 Forbidden / X-Proxy-Error: bl
CLAUDE.md · 68 lines

How it starts

The opening of the file, as written. The whole thing — 68 lines — stays where its author put it; the contents beside it link to each section on GitHub.

CLAUDE.md — claude-plugins

Developer notes for the MSApps Claude Plugins repository.


Repo Structure

plugins/
  {plugin-name}/
    README.md
    CONNECTORS.md
    skills/{skill-name}/
      SKILL.md
      references/
    commands/
      {command}.md

Known Issues & Fixes

session-backup: VM Proxy Blocks Google Drive Uploads

Problem: The Cowork VM proxy blocks outbound connections to script.google.com. Running curl or requests from inside the VM to the Google Apps Script upload endpoint fails with 403 Forbidden / X-Proxy-Error: blocked-by-allowlist.

Fix (as of 2026-03-27): Upload must run on the Mac side via Desktop Commander:

  1. Copy zip files from /tmp/cowork-backup/ to {SESSION_MNT}/Claude/.backup-tmp/ (mounted Mac folder)
  2. Write a Python upload script to the staging dir
  3. Use mcp__Desktop_Commander__start_process to run python3 ~/Documents/Claude/.backup-tmp/do_upload.py on the Mac
  4. Clean up staging dir after upload

See plugins/session-backup/skills/session-backup/SKILL.md Step 4 for full implementation.

session-backup: Hardcoded Session Paths Break Between Runs

Problem: Cowork VM session names (e.g. loving-gracious-allen, modest-sleepy-carson) are randomly generated and change with every new session. Any hardcoded path like /sessions/loving-gracious-allen/mnt/... will fail in a different session.

Fix (as of 2026-03-27): Always discover the active session dynamically:

import glob
candidates = glob.glob('/sessions/*/mnt/Claude/credentials/cowork-gdrive-config.json')
session_mnt = candidates[0].replace('/Claude/credentials/cowork-gdrive-config.json', '')

Rule: Never hardcode a session name in any skill or scheduled task prompt.


Deployment

Plugins are published via:

  • GitHub: MSApps-Mobile/claude-plugins (this repo)
  • Marketplace: Listed in Anthropic Claude plugin marketplace

To install from this repo in Cowork:

Settings → Plugins → Marketplaces → Add → MSApps-Mobile/claude-plugins

Read the full file on GitHub · 68 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. 3d ago First seen · 68 lines · 556 tokens per session scan C 5b5788e68105

Subscribe to this mod's changes

claude-plugins CLAUDE.md is an instructions file published in the GitHub repository MSApps-Mobile/claude-plugins (9 stars, last pushed 7d ago), licensed MIT. It adds 556 tokens to every session, about $0.0028 per session on Opus 5. A static security scan graded it C with 2 findings (enumerates the file system for secrets, makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.