opencode--skilled-agent-loops-with-spec-kit-memory

Routes non-trivial requests to matching skills through standalone MCP metadata and stable advisor tool ids.

Unified spec-folder workflow + context preservation: Levels 1-3+, validation, Spec Kit Memory. Required for file modifications.

OpenCode CLI orchestrator: external dispatch, in-OpenCode parallel sessions, cross-AI handback with full runtime context.

Shared deep-loop runtime: executor + prompt-pack + validation + atomic state + coverage-graph + Bayesian scoring + fallback routing.

Iterative codebase-context-gathering deep loop. Runs a configurable pool over a shared scope in parallel (native-only by default; optional heterogeneous CLI seats) and synthesizes a reuse-first Context Report for planning/implementation. Use before /speckit:plan or /speckit:implement to map existing code, integration points, and conventions.

Unified deep-loop workflow skill: routes a request to one of five modes (context, research, review, ai-council, improvement) over the shared deep-loop-runtime backend. Holds no per-mode logic — it dispatches by workflowMode through mode-registry.json. Use for codebase-context gathering, autonomous research, iterative code review, multi-seat AI Council planning, and evaluator-first agent/model/skill/non-dev-system improvement.

MCP orchestration via TypeScript execution. Use Code Mode for ALL external MCP tool calls; ~98% context reduction, type-safe.

Drive the installed Open Design desktop app from the terminal through its `od` CLI and stdio MCP server. Read local design projects and design-systems, reuse them, answer the app's prompts, and commission headless generation runs without using the in-app chat. Wires Open Design's MCP server into opencode or Claude Code. MANDATORY: any UI/design work through Open Design also requires sk-interface-design — this skill is the transport, that skill is the non-negotiable design judgment.

Claude Code CLI executor for Anthropic-backed reasoning, edits, reviews, and structured cross-AI handoff.

Codex CLI executor for OpenAI-backed coding, repo analysis, PR review, web research, and cross-model validation.

AI Council: multi-seat planning, artifact persistence, convergence checks, packet-local ai-council outputs.

Evaluator-first bounded agent improvement: 5-dim scoring, dynamic profiling, packet-local candidates, guarded promotion.

Autonomous deep-research loop: iterative investigation, externalized state, convergence detection, fresh context per pass.

Autonomous iterative code-review loop with externalized state, convergence detection, P0/P1/P2 findings, fresh context per pass.

Chrome DevTools orchestrator: routes between bdg CLI (fast, token-efficient) and Code Mode MCP (multi-tool integration).

Routes ClickUp between cupt CLI (daily ops) and official MCP (docs, goals, bulk). Embedded install and agent safety invariants.

Figma CLI orchestrator: drives figma-ds-cli for terminal Figma design work, with an optional Figma MCP via Code Mode.

Multi-stack coding standards and verification. Smart router auto-detects the active surface and loads matching code patterns.

Markdown and OpenCode component specialist: doc quality, content optimization, component scaffolds, install guides, playbooks.

Git workflow orchestrator OWNING all git/version-control intent: git worktree create/restructure (numbered wt/{NNNN}-{name} branches under .worktrees/), branch, conventional commits, pull request (PR), merge, rebase, finish work, integrate changes. Routes git-worktrees, git-commit, git-finish. NOT for spec folders / memory / save context (system-spec-kit) or code implementation / tests (sk-code).

Prompt engineering specialist: structured AI prompts via 7 frameworks, DEPTH thinking, CLEAR scoring.

Per-model prompt-craft hub for small-model dispatch (DeepSeek-v4-pro + Kimi-k2.7-code + MiniMax-M3 + MiMo-V2.5-Pro via cli-opencode). OWNS the per-model prompt-craft profiles in references/models/ (framework + scaffold + gotchas, mirroring model_profiles.json); executor MECHANICS (binary flags, invocation wrappers) stay in cli-opencode. Advisor co-surfaces it with that executor.

Structural code indexing and mk-code-index MCP workflows for graph readiness, impact queries, context retrieval and structural checks.

Stack-agnostic code-review baseline: findings-first severity, mandatory security/correctness minimums, sk-code evidence.

Guidance for distinctive, intentional UI design when building or reshaping an interface. Drives deliberate palette, typography, layout and motion choices that avoid templated AI defaults, with a brainstorm-critique-build process and interface writing rules.

Expert code review of current git changes with a senior engineer lens. Detects SOLID violations, security risks, and proposes actionable improvements.