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.
git 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/commands/nyldn/claude-octopus/octo-deliver)<a href="https://agentmods.dev/commands/nyldn/claude-octopus/octo-deliver"><img src="https://agentmods.dev/badge/commands/nyldn/claude-octopus/octo-deliver/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/commands/nyldn/claude-octopus/octo-deliver"><img src="https://agentmods.dev/badge/commands/nyldn/claude-octopus/octo-deliver.svg" alt="Reviewed on agentmods" width="80" 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 | $0.00015 | $0.00998 |
| Opus 5 | $0.00008 | $0.00499 |
| Sonnet 5 | $0.00003 | $0.00200 |
| Haiku 4.5 | $0.00002 | $0.00100 |
Grade A, and why
octo-deliver 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 8d 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.
Copies of this mod
2 near-identical copies found in the catalogue:
- deliver — 97% identical, 6 lines differ
- octo-deliver — 88% identical, 13 lines differ
How it starts
The opening of the file, as written. The whole thing — 123 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Deliver - Delivery Phase ✅
🤖 INSTRUCTIONS FOR CLAUDE
MANDATORY COMPLIANCE — DO NOT SKIP
When the user explicitly invokes /octo:deliver, you MUST execute the structured workflow below. You are PROHIBITED from doing the task directly, skipping the validation/review phase, or deciding the task is "too simple" for this workflow. The user chose this command deliberately — respect that choice.
EXECUTION MECHANISM — NON-NEGOTIABLE
You MUST execute this command by reading and following the corresponding workflow source. You are PROHIBITED from:
- ❌ Using the Agent tool to research/implement yourself instead of invoking the skill
- ❌ Using WebFetch/Read/Grep as a substitute for multi-provider dispatch
- ❌ Skipping
orchestrate.shcalls because "I can do this faster directly" - ❌ Implementing the task using only Claude-native tools (Agent, Write, Edit)
Multi-LLM orchestration is the purpose of this command. If you execute using only Claude, you've violated the command's contract.
When the user invokes this command (e.g., /octo:deliver <arguments>):
✓ CORRECT - Read the explicit workflow source:
Read ${HOME}/.claude-octopus/plugin/.claude/skills/flow-deliver/SKILL.md, then execute it with <user's arguments>
✗ INCORRECT:
Skill(skill: "flow-deliver", ...) ❌ Wrong! Octopus skills are model-invocation disabled
Task(subagent_type: "octo:deliver", ...) ❌ Wrong! This is a skill, not an agent type
Auto Code Review & E2E Verification (MANDATORY)
Before presenting results, launch two verification agents in parallel:
Agent(model: "sonnet", subagent_type: "feature-dev:code-reviewer", run_in_background: true,
description: "Code review: deliver phase",
prompt: "Review code changes from this session. Check git diff. Report only high-confidence bugs, security issues, and convention violations.")
Agent(model: "sonnet", run_in_background: true,
description: "E2E test: deliver phase",
prompt: "Run the project's test suite. Report tests passed/failed and any regressions.")
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.
- 8d ago First seen · 123 lines · 15 tokens per session scan A d088bc2dccf8
octo-deliver is a command published in the GitHub repository nyldn/claude-octopus (4,061 stars, last pushed yesterday), licensed MIT. It adds 15 tokens to every session and 998 once invoked, about $0.0001 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 commands, from other repositories
tdd
Test-driven development with tests that actually catch breaks: red before green, name the break each test catches, and the mutation check before done.
test
Run the repository's actual test suite: every ecosystem's canonical runner — NOT run is never green.
v-collect
Re-run the collect + scope-gate + review tail of a Compound V run by run-id. Normalizes each job's output into results/ .json, re-runs the git-derived scope gate on every job, then the three-pass Review Gate (AC-gated). Use to re-check a dispatched run without re-dispatching workers.
ia-test-browser
Run browser tests on pages affected by current PR or branch.
ia-verify
Run pre-PR verification chain -- build, types, lint, tests, security scan, diff review.
evaluator
You are the Evaluator: a precision tool for assessing project work, code, and strategy. You hold a high bar and give unsparing, evidence-based verdicts — with full respect for context, test status, and agreed tradeoffs.