OpenRig is a software harness that coordinates Claude Code and Codex coding agents as one system, with agent teams defined in YAML and launched together. It is for organizing persistent teams of coding agents that can delegate work to specialists and return results or decisions. The catalogue lists 53 skills and one plugin as add-ons for OpenRig.
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 skills add mvschwarz/openrig --skill agent-operated-workflowsgit clone --depth 1 https://github.com/mvschwarz/openrigWrote 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/skills/mvschwarz/openrig/agent-operated-workflows)<a href="https://agentmods.dev/skills/mvschwarz/openrig/agent-operated-workflows"><img src="https://agentmods.dev/badge/skills/mvschwarz/openrig/agent-operated-workflows/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/skills/mvschwarz/openrig/agent-operated-workflows"><img src="https://agentmods.dev/badge/skills/mvschwarz/openrig/agent-operated-workflows.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector pass
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 | $0.00055 | $0.00844 |
| Opus 5.5 | $0.00022 | $0.00338 |
| Sonnet 5.5 | $0.00011 | $0.00169 |
| Haiku 4.5 | $0.00006 | $0.00084 |
Grade A, and why
agent-operated-workflows 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 1mo 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.
How it starts
The opening of the file, as written. The whole thing — 82 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Agent-Operated Workflows
An Agent-Operated Workflow is a bounded procedure whose control loop belongs to an agent. The agent observes, interprets, sequences bounded actions, verifies their effects, adapts, and escalates genuine ambiguity. Deterministic code remains the exact substrate; it does not pretend to understand every state of a brownfield world.
Name the architecture before using it
- AI-enabled software: application code owns the control loop and calls a model as one capability.
- Agent-Operated Workflow: an agent owns one bounded procedure with a start and stop condition.
- Agent-Operated Software: an OpenRig agent or rig participates in an ongoing application's live
backend or control loop. It may run many workflows; see
agent-operated-software.
Calling a model is not enough. A deterministic program that consults a model still owns its loop.
Decide whether the agent earns the loop
Use an agent when the happy path is deterministic but the environment is brownfield, variable, partially knowable, or disproportionately expensive to model exhaustively. The judgment is in discovering what is true and choosing the next safe act.
Prefer ordinary code when the state space is closed and code is simpler and more reliable than judgment. A parser, fixed data transformation, or fully enumerated state machine does not become better because an agent drives it.
Keep ownership explicit
| Owner | Responsibility |
|---|---|
| Runbook | owns policy, the mental model, checkpoints, stop conditions, and escalation rules |
| Deterministic tools | own context gathering and bounded exact actions |
| Agent | owns interpretation, sequencing, adaptation, and effect verification |
| Human | owns destructive ambiguity and product policy |
The runbook guides judgment; tools enforce the invariants that must not be arguable.
Operate
- Establish identity, durable state, the runbook, and the promised outcome.
- Inspect with side-effect-free reads or an explicit plan mode.
- Choose one bounded action from observed state; do not mechanically replay the happy path.
- Verify the effect at the consumer or state surface that could prove it wrong.
- Record an observable checkpoint before moving on, so another agent can resume.
- Continue, stop at the declared condition, or escalate with evidence and the decision needed.
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.
- 1mo ago Changed · +2 lines 24e54cbfa979
- 1mo ago First seen · 80 lines · 55 tokens per session scan A 0d52c34f6144
agent-operated-workflows is a skill published in the GitHub repository mvschwarz/openrig (7,515 stars, last pushed today), licensed Apache-2.0. It adds 55 tokens to every session and 844 once invoked, about $0.0002 per session on Opus 5.5. 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 2026-09-05.
Other skills, from other repositories
metis
A task-routing and checking workflow for explicitly requested Metis work. It assigns work to defined roles, records checks, and limits recovery when something fails.
metis-browser
Control Metis Desktop Inspector built-in browser for local preview (HTML/SVG/file paths), page interaction, accessibility snapshots, and screenshots. Use when the user asks to open, show, preview, inspect, click, type, or verify UI in the in-app/built-in browser. Prefer browsernavigate over bash…
dynamic-resources
Example skill loaded from resourcesdiscover.
browser-use
Control the user's Chrome through the existing persistent Node REPL and the Qwen Browser SDK.
verify
Exercise the real app/API/CLI and collect observable evidence; tests alone do not count as end-to-end verification.
swarm-pr-review
Run a graph-guided, tool-augmented PR review using context packing, parallel exploration, mandatory repository-agnostic risk-family coverage with dispatch scaled to diff size and risk, independent reviewer validation, critic challenge, and metrics writeback. Use for deep pull request review with low false-positive…