Skills for Real Engineers. Straight from my .agents directory.
agent-operated-workflows
Agent skill frommvschwarz/openrig
Use when a bounded real-world procedure has a deterministic happy path but brownfield, variable, or partially knowable state; when an operation must resume from verified evidence; or when deciding whether agent judgment or ordinary code should own a procedure's control loop.
How to install agent-operated-workflows
npx skills add mvschwarz/openrig# or add only this skill to Claude Code: copy the folder
packages/daemon/assets/plugins/openrig-core/skills/agent-operated-workflows/
# into
~/.claude/skills/agent-operated-workflows/ (all projects)
.claude/skills/agent-operated-workflows/ (one project)
packages/daemon/assets/plugins/openrig-core/skills/agent-operated-workflows/SKILL.md
Apache-2.0Full file1# Agent-Operated Workflows23An **Agent-Operated Workflow** is a bounded procedure whose control loop belongs to an agent. The4agent observes, interprets, sequences bounded actions, verifies their effects, adapts, and escalates5genuine ambiguity. Deterministic code remains the exact substrate; it does not pretend to understand6every state of a brownfield world.78## Name the architecture before using it910- **AI-enabled software:** application code owns the control loop and calls a model as one capability.11- **Agent-Operated Workflow:** an agent owns one bounded procedure with a start and stop condition.12- **Agent-Operated Software:** an OpenRig agent or rig participates in an ongoing application's live13 backend or control loop. It may run many workflows; see `agent-operated-software`.1415Calling a model is not enough. A deterministic program that consults a model still owns its loop.1617## Decide whether the agent earns the loop1819Use an agent when the happy path is deterministic but the environment is brownfield, variable,20partially knowable, or disproportionately expensive to model exhaustively. The judgment is in21discovering what is true and choosing the next safe act.2223Prefer ordinary code when the state space is closed and code is simpler and more reliable than24judgment. A parser, fixed data transformation, or fully enumerated state machine does not become25better because an agent drives it.2627## Keep ownership explicit2829| Owner | Responsibility |30|---|---|31…
Excerpt shown under the repository's Apache-2.0 license.
More skills in openrig
Similar skill repositories
An agentic skills framework & software development methodology that works.
A single CLAUDE.md file to improve Claude Code behavior, derived from Andrej Karpathy's observations on LLM coding pitfalls.
Public repository for Agent Skills
Makes your AI agent think like the laziest senior dev in the room. The best code is the code you never wrote.
Persistent Context Across Sessions for Every Agent – Captures everything your agent does during sessions, compresses it with AI, and injects relevant context back into future sessions. Works with Claude Code, OpenClaw, Codex, Gemini, Hermes, Copilot, OpenCode + More
