Claude Octopus is an orchestration project that sends research, design, and coding tasks to Claude Code and other AI model providers so their results can be compared. Developers use it for multi-model work, disagreement detection, reviews, persistent context, and an optional workflow that moves from discovery through delivery. The catalogue entries are its commands, skills, agents, instructions, hooks, plugins, and settings.
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/nyldn/claude-octopus/agents-mdgit clone --depth 1 https://github.com/nyldn/claude-octopusWrote 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/nyldn/claude-octopus/agents-md)<a href="https://agentmods.dev/instructions/nyldn/claude-octopus/agents-md"><img src="https://agentmods.dev/badge/instructions/nyldn/claude-octopus/agents-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 | $0.01524 | $0.01524 |
| Opus 5 | $0.00762 | $0.00762 |
| Sonnet 5 | $0.00305 | $0.00305 |
| Haiku 4.5 | $0.00152 | $0.00152 |
Grade A, and why
claude-octopus AGENTS.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 yesterday.
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 — 120 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Claude Octopus Agents
This file describes the autonomous agents available in this repository for AI coding tools that support agent discovery (e.g., GitHub Copilot coding agent).
Available Agents
| Agent | Description | Tools |
|---|---|---|
backend-architect |
Scalable API design, microservices, distributed systems | Read-only |
code-reviewer |
Code quality, security vulnerabilities, production reliability | Read-only |
debugger |
Errors, test failures, unexpected behavior | All |
docs-architect |
Technical documentation from codebases | Read + execute |
frontend-developer |
React components, responsive layouts, client-side state | All |
performance-engineer |
Optimization, observability, scalable performance | Read-only |
security-auditor |
DevSecOps, OWASP compliance, vulnerability assessment | Read-only |
tdd-orchestrator |
Red-green-refactor discipline, test-driven development | All |
database-architect |
Data modeling, schema design, migration planning | Read-only |
cloud-architect |
AWS/Azure/GCP infrastructure, IaC, FinOps | Read + execute |
Agent Definitions
Agents are defined in two formats for cross-platform compatibility:
- Claude Code:
.claude/agents/*.md— YAML frontmatter with Claude Code tool names - GitHub Copilot:
.github/agents/*.agent.md— YAML frontmatter with Copilot tool aliases
Both formats describe the same 10 agents with platform-native tool mappings:
| Claude Code Tool | Copilot Alias |
|---|---|
| Read | read |
| Write, Edit | edit |
| Bash | execute |
| Grep, Glob | search |
MCP Integration
The MCP server (mcp-server/) exposes Claude Octopus workflows as MCP tools.
For MCP-aware coding agents, connect to the MCP server rather than invoking
agents directly.
Repo Orientation for Agents
Mirror of the "Repo Orientation for Agents" section in CLAUDE.md (keep both in sync). Essentials:
- Derived artifacts, never hand-edit:
.claude-plugin/marketplace.json(regen:./scripts/sync-marketplace.sh) andopenclaw/src/tools/index.ts(regen:./scripts/build-openclaw.sh).scripts/sync-readme.pyowns current facts across both READMEs andPRODUCT.md. Runmake syncafter component, metadata, release, model, or provider changes. During development use focused suites; before an ordinary branch push runmake ci-changed, which fails closed to the full matrix for shared or unmapped changes. Runmake ci-localbefore merge and release. - Exec bits: scripts stay
100755;git diff origin/main...HEAD --summary | grep "mode change"must be empty (CI-enforced;allow-mode-changelabel bypasses). Local test runs chmod test fixtures as a side effect — recheck modes after every local test run. - Provider wiring: 7-point checklist across 5 files, documented in
docs/PROVIDERS.md. Case globs are order-sensitive (claude-sdk*beforeclaude*). - Releases: follow
RELEASING.md; tag the squash-merge commit onmain, never the branch head. - Secret-scan quoting: write
"SOME_API_KEY=${VAR}", notSOME_API_KEY="${VAR}". - Outbound GitHub text: never place generated text or Markdown directly in shell
--body,-f body=, or similar arguments. Stream it to./scripts/safe-gh-comment.sh --repo OWNER/REPO ... -on stdin (or pass a private file); the helper snapshots and validates the body before a silent GitHub write. - beads blocked? If bd writes are blocked by pending schema migrations, do NOT migrate; record work in the session handoff and flag the blockage.
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.
- yesterday First seen · 120 lines · 1,524 tokens per session scan A c815c736c164
claude-octopus AGENTS.md is an instructions file published in the GitHub repository nyldn/claude-octopus (4,035 stars, last pushed yesterday), licensed MIT. It adds 1,524 tokens to every session, about $0.0076 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 2026-09-03.
Other instructions, from other repositories
platform-skills CLAUDE.md
Claude Code instructions for nitinjain999/platform-skills, covering platform skills - skill development guide, philosophy, production-first engineering, progressive disclosure and multi-domain coherence.
contextdocs AGENTS.md
Instructions for littlebearapps/contextdocs, covering contextdocs, identity, agent, available skills and workflow commands.
soleur AGENTS.md
AGENTS.md instructions for jikig-ai/soleur, covering agent instructions — index, hard rules, workflow gates, code quality and review & feedback.
claude-setup CLAUDE.md
Instructions for shahtuyakov/claude-setup, covering claude.md, repository overview, structure, knowledge base and skill architecture.
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).
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.