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/web3dev1337/agent-workspace/claude-mdgit clone --depth 1 https://github.com/web3dev1337/agent-workspaceWhat 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 | $0.11378 | $0.11378 |
| Opus 5 | $0.05689 | $0.05689 |
| Sonnet 5 | $0.02276 | $0.02276 |
| Haiku 4.5 | $0.01138 | $0.01138 |
Grade F, and why
agent-workspace CLAUDE.md scanned grade F with 6 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Sends data to an external URLmediumData exfiltration
A POST to an outside endpoint may be telemetry or may be exfiltration; either way the mod talks to somewhere, and you should know where.
`curl -sS -X PATCH http://localhost:$PORT/api/commander/instances/$COMMANDER_INSTANCE_ID -H 'Content-Type: application/json' -d '{"label":"Deploys"}'` Asks for rootmediumPrivilege escalation
A mod that escalates privileges can change anything on the machine, not only the project.
-d '{"package":"dos2unix","reason":"CRLF fix for WSL scripts","installCmd":"sudo apt-get install -y dos2unix","category":"apt"}' Downloads and executes remote codehighSupply chain
curl | sh runs whatever the server returns today, which is not necessarily what it returned when this was reviewed.
curl --proto '=https' --tlsv1.2 -sSf https://sh.rustup.rs | sh -s -- -y && source "$HOME/.cargo/env" Reads agent configuration directoriesmediumAgent snooping
.claude/, .codex/, .gemini/ hold keys, settings and other credentials a mod has no legitimate need for.
- Config: `~/.codex/config.toml` Recursive force deletehighDestructive command
rm -rf with a variable or a broad path is one typo away from removing the wrong tree.
rm -rf ~/.nvm/versions/node/v24.9.0/lib/node_modules/@openai/codex && npm i -g @openai/codex@latest Makes network callslowCapability
Not a fault in itself. Listed so you know the mod talks to something, and to what.
curl -sS "https://api.trello.com/1/cards/CARD_ID/customFieldItems?key=$KEY&token=$TOKEN" | jq -r '.[0].idValue' How it starts
The opening of the file, as written. The whole thing — 963 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Claude Orchestrator Guidelines for Claude Code
🚨 READ THIS ENTIRE FILE 🚨 CRITICAL: You MUST read this complete file from start to finish. Do not truncate or skip sections.
Note: This is a revolutionary multi-workspace orchestrator for managing unlimited Claude Code sessions with mixed-repository support, dynamic worktree creation, and zero-friction workflows.
🚨 IMPORTANT: ALWAYS PROVIDE PR URL 🚨
When creating any pull request, ALWAYS provide the PR URL in your response to the user. This is mandatory for all PRs.
Resume / Context Reset (read this if you lose context)
If you ever lose context mid-run, do not improvise. Resume from these files:
PLANS/2026-01-20/REQUESTED_CHANGES.md(master list of requested fixes)PLANS/2026-01-20/ROLLING_LOG.md(what shipped; PR links; where to resume)PLANS/2026-01-24/OPTIMAL_ORCHESTRATOR_PROCESS.md(tier + queue model)PLANS/2026-01-25/WORKFLOW_TIER_RISK_PROMPTS.md(current open items: tiers, risk, prompts, review inbox)PLANS/2026-01-25/BRAIN_DUMP_2026-01-25.md(full workflow brain dump transcript)PLANS/2026-01-25/BRAIN_DUMP_IMPLEMENTATION_PLAN.md(what’s missing + next PRs)PLANS/2026-01-25/DATA_MODEL.md(where tiers/deps/prompts live)
Current open PRs (keep updated in the rolling log):
- None (check with
gh pr list --state open)
Process workflow controls (Commander + voice)
Commander/voice can drive the tiered workflow via semantic commands:
set-workflow-mode(focus|review|background)set-focus-tier2(auto|always)open-queue,open-tasks,open-advice
Tier tagging persistence:
- stored in
~/.orchestrator/task-records.jsonviasession:<id>task records - loaded on page refresh via
GET /api/process/task-records
Launching Agents from Trello
Board-to-Repo Mapping
CRITICAL: Always check ~/.claude/TRELLO_BOARDS.md for the full board/list/repo mapping. Do NOT assume a board maps to a specific repo - look it up!
| Board | Board ID | Repo Path | Repo Type |
|---|---|---|---|
| Zoo Hytopia | 691e5516c77f3e9c9fd89f61 |
~/GitHub/games/hytopia/zoo-game/ |
hytopia-game |
| Arcade World | 694a07bae349c125d4568094 |
~/GitHub/games/hytopia/games/hytopia-2d-game-test/ |
hytopia-game |
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 · 963 lines · 11,378 tokens per session scan F 5e9ad2b3af3c
agent-workspace CLAUDE.md is an instructions file published in the GitHub repository web3dev1337/agent-workspace (33 stars, last pushed 2d ago), licensed MIT. It adds 11,378 tokens to every session, about $0.0569 per session on Opus 5. A static security scan graded it F with 6 findings (sends data to an external url, asks for root, downloads and executes remote code). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other instructions, from other repositories
ii-agent AGENTS.md
Instructions for Intelligent-Internet/ii-agent, covering ii-agent contributor guide, quick start, repository map, mandatory rules and architecture.
intelligent-terminal rust.instructions.md
Concise Rust coding conventions for this repository.
Browser4 CLAUDE.md
Instructions for platonai/Browser4, covering browser4 — project context for claude, architecture, key dispatch chain (cli → browser), batch commands and e2e test structure.
nimbalyst AGENTS.md
AGENTS.md instructions for nimbalyst/nimbalyst: At the beginning of every AI session, read the root CLAUDE.md and any applicable package-level CLAUDE.md before taking actions.
sortie copilot-instructions.md
Instructions for sortie-ai/sortie, covering sortie coding & review standards, 1. layered imports (downward only; violation is critical), 2. concurrency safety, 3. workspace path safety (critical: security boundary) and 4. persistence (sqlite).
GENesis-AGI AGENTS.md
AGENTS.md instructions for WingedGuardian/GENesis-AGI, covering agent instructions, code review mandate (adversarial), gitnexus — code intelligence (advisory), genesis capability surface and skills.