Menú
Use when a Mister Smith phase, issue, PR, or design decision needs a frontier-legitimacy judgment about supervised autonomy, scope drift, or security-versus-autonomy balance.
Instalar con Codex o Claude Copia este prompt, pégalo en Codex, Claude u otro asistente, y deja que revise la página de la skill y la instale por ti.
Basado en la clasificación ocupacional SOC
Use when any Mister Smith request touches Linear, GitHub PR flow, packet slicing, review state, workspace hygiene, or bootstrap readiness and needs the correct Smith MCP route first.
Use after a Mister Smith SpecKit packet has been created and needs post-spec closure: packet verification, git closure, Linear materialization, honest execution-lane preparation, and follow-on execution readiness without implementation unless explicitly requested.
Use when the user wants a complete Mister Smith live runtime session or workflow evaluation with real execution, durable evidence capture, and honest task/session/autonomy proof instead of mock-only validation.
Use when Mister Smith needs spec-only frontier packet work: turning current repo truth or a pre-SpecKit primer into a new product-side SpecKit packet, or running the full local SpecKit authoring chain for an already chosen packet. Trigger for requests to create or extend a Mister Smith packet spec, pre-SpecKit primer, specify-init prompt, or full spec bundle without implementation.
Use when a Mister Smith SpecKit phase or tasks pack needs to be turned into deterministic Linear slices, blocker chains, prep-slice opportunities, and direct-execution-ready work.
Legacy-named skill for Mister Smith tasks that span Linear control-plane state, GitHub PR flow, local repo truth, lifecycle recovery, or workspace hygiene.
| name | mister-smith-frontier-mandate |
| description | Use when a Mister Smith phase, issue, PR, or design decision needs a frontier-legitimacy judgment about supervised autonomy, scope drift, or security-versus-autonomy balance. |
Mister Smith is a first-class multi-agent orchestration operating system in Rust, built on NATS/JetStream and Erlang OTP-inspired supervision trees. It is not being built to follow the current agent-framework market. It is being built to define the standard that market will later converge toward.
Do not choose an approach because it is popular, familiar, or already normalized by OpenAI Agents SDK, Google ADK, LangChain/LangGraph, CrewAI, AutoGen, Claude SDK, or similar systems. Benchmark them. Learn from them. Then exceed them. Pull from distributed systems, actor systems, operating systems, telecom, trading infrastructure, and real-time messaging when those fields offer stronger patterns.
Reuse what is already correct. Do not reinvent primitives without benefit. But wherever the choice affects Mister Smith’s core advantage—coordination, execution, supervision, memory, streaming, routing, reliability, observability, state, or distributed behavior—prefer the architecture with the highest long-term leverage, scalability, and strategic value, even when it is less conventional.
Incremental imitation is failure. Favor well-reasoned designs that create real advantage. Recommend unconventional or experimental methods when they are materially superior. Build with the standard-setting mindset of a team creating the framework others will later copy.
Use the smith MCP tools first for frontier-legitimacy judgments.
evaluate_issue_legitimacyclassify_follow_up_work