Explore

Apply Shadmin feature-development standards (backend Go/Gin/Ent + frontend React/TS). Use when adding/modifying features, CRUD modules, API routes/controllers/usecases/repositories, Ent schemas, frontend pages/routes, React components, TanStack hooks, or any full-stack work in this project. Trigger whenever the user mentions new features, backend changes, frontend changes, database schema changes, permissions, UI pages, tables, forms, or API endpoints — even if they don't explicitly say "feature development."

Safe cleanup of Shadmin's Go backend, React frontend, Go CLI, and docs — remove dead code, consolidate duplication, and improve maintainability without changing behavior or adding features. Use when asked to simplify, deduplicate, remove dead code, tidy up, refactor for maintainability, or align naming and conventions.

Reconcile and maintain useful personal and project memory with the native memory tools.

Use when diagnosing a evener session, job, watch, or the session tree — investigating watch runaways (self-influence depth + runaway-fuse drops), dropped/coalesced deliveries, "how many times did X actually get called", stuck turns, or parent↔delegate↔observer linkage. Covers the evener doctor forensic tools, the Finding contract, runbook authoring, and graduated repair.

Use when designing, running, or diagnosing Evener tool-fluency experiments for a real built-in, plugin, or newly added tool. Covers probe manifests, model comparisons, semantic oracles, and interpreting fluency failures without ad-hoc transcript parsing.

Use when a surface parses untrusted or model-generated input — a wire protocol, tool-call JSON, upstream SSE, an HTTP body — or after a parser/dispatch bug, to fuzz that surface and turn any deterministic failure into a permanent, flake-guarded regression test. Covers picking a seam, choosing oracles beyond "no panic", the portable promoter/schema-generator core, and the safety bounds that keep a fuzzer from executing dangerous handlers.

Run when the user invokes /rpa-init or asks to onboard or warm up context on a repository. The agent studies code, reads documentation and test code, sets up the dev environment as the project expects, runs tests, and writes a short project report. No extra user brief is required.

Use when the user invokes /rpa-gen-rules or asks to create, refresh, audit, or synchronize project instructions such as root AGENTS.md, Cursor .cursor/rules, or Claude Code .claude/rules, especially when existing rule trees have drifted.

Run when the user invokes /rpa-feat together with a clear feature description (for example issue text). BDD workflow: plan, failing tests, implementation, green tests, full suite, docs and examples, linter at the end. Will not work correctly without the user supplying what to build.

Run when the user invokes /rpa-bugfix together with a clear bug description (expected vs actual, reproduction). Reproduction test first, then fix, full test suite, short report. Will not work correctly without the bug details.

Change Coddy's own configuration when the user asks for it: edit settings, providers, models, logging, permissions, or find, install, update, and remove MCP servers and skills. Stages UCI-style commands for config.yaml and commits only after the user confirms saving; MCP servers are entries of the mcp.json files. Load when the user explicitly asks to change a Coddy setting, or when the request implies it (install an MCP server, add a skill, switch a model, roll back the config). Do not load for ordinary coding or unrelated tasks.

Add descriptions for new models from the HuggingFace router to chat-ui configuration. Use when new models are released on the router and need descriptions added to prod.yaml and dev.yaml. Triggers on requests like "add new model descriptions", "update models from router", "sync models", or when explicitly invoking /add-model-descriptions.

Run a workflow — drive an MCP workflow lifecycle (execute/pause/resume/cancel) or invoke + resume a native .claude/workflows/*.js orchestration via the Workflow tool

Author a workflow — either an MCP workflow template (persisted, lifecycle) or a native .claude/workflows/*.js orchestration script (agent/parallel/pipeline fan-out)

Walk through a complete GAIA benchmark→submit flow — from key resolution through HAL-compatible package generation

Diagnose why a GAIA question failed — extract trace, classify failure mode, and propose a fix. Use when a GAIA benchmark run reports a failed/incorrect task_id and you need to root-cause it before resubmitting.

Side-by-side comparison of ruflo vs HAL vs other GAIA harnesses — capability gaps, design decisions, and improvement roadmap

Detect missing test coverage and generate test suggestions. Use when the user asks about coverage gaps, untested code, or what tests to write next; also after adding a feature to find what still needs tests.

Test-Driven Repair — given a failing test, spawn a bounded headless `claude -p` (Read/Edit/Bash only) that makes the test pass without modifying it. Modeled on agent-harness-generator's ADR-175 Test-Driven Repair mode. Bounded cost via --max-budget-usd, bounded capability via --allowedTools. Closes the loop the TDD plugins didn't — we generate tests, this fixes the code to satisfy them.

Initialize a multi-agent swarm with anti-drift configuration. Use when starting a complex multi-file task that needs 3+ coordinated agents (feature implementation, refactor across modules, security audit). Skip for single-file edits or quick questions.

Stream live swarm events using the Monitor tool for real-time observability

Run the SPARC Specification phase — gather requirements, define acceptance criteria, identify constraints, and store the spec in memory

Run the SPARC Refinement and Completion phases — review code, improve test coverage, validate against specification, and generate documentation

Run the SPARC Pseudocode and Architecture phases (2 and 3) — write algorithm pseudocode, design module boundaries and API contracts, then implement

Run full security scans on the codebase using Ruflo security tools. Use when reviewing PRs for security regressions, auditing auth/input-handling code, before production deploys, or when the user asks for a security check at quick/standard/deep depth.

Scan project dependencies for known vulnerabilities and CVEs. Use when auditing third-party packages, before releases, after `npm install`/lockfile changes, or when investigating reported CVE advisories.

Persist and restore agent sessions across conversations with state snapshots

Manage RVF (Ruflo Vector Format) files for portable agent memory and cross-platform transfer

Format chat messages in a local model's template (llama3, mistral, chatml, phi, gemma, or a model id) and route queries over a small WASM HNSW pattern index (about 11 patterns). Use when preparing prompts for local runtimes.

First-run setup for ruvector@0.2.25 — installs ONNX/Brain/SONA add-ons, registers the MCP server, and verifies the install via `doctor`

Embed hierarchical data via npx ruvector@0.2.25 embed text and project into the Poincare ball in user code (no --model poincare flag in 0.2.25)

Generate embeddings via npx ruvector@0.2.25 embed text (ONNX all-MiniLM-L6-v2, 384-dim), normalize, and store in HNSW index

Cluster code by graph community detection via npx ruvector@0.2.25 hooks graph-cluster (spectral / Louvain)

Run a ruflo swarm agent on the user's own ruOS cloud desktop through the tenant-authenticated fleet MCP, stream its output, and record it in ruflo's swarm state. Use when the user wants an agent to run on a ruOS desktop, asks to "run this on my ruOS desktop", or places a swarm agent on a remote ruOS host.

SOTA semantic search — hybrid (sparse+dense), Graph RAG multi-hop, MMR diversity reranking, recency weighting

Bridge Claude Code auto-memory into AgentDB with ONNX embeddings, deduplicate, and enable unified cross-project search

Validate a Claude Code plugin structure, frontmatter, and MCP tool references

Scaffold a new Claude Code plugin with proper directory structure, plugin.json, skills, commands, and agents

Trace agent execution by collecting spans and building a trace tree for a task

Aggregate and display system metrics with anomaly detection for a time period

Train neural models (LSTM, Transformer, N-BEATS) on market data using npx neural-trader with confidence intervals

Generate trading signals using npx neural-trader anomaly detection engine with Z-score scoring and neural prediction

Assess portfolio risk using npx neural-trader — VaR, CVaR, Sharpe, position sizing, circuit breaker status

Detect current market regime using npx neural-trader — bull/bear/ranging/volatile classification with recommended strategy. Use when the user asks about market conditions, wants to pick a strategy for current conditions, or before running a backtest/signal that should be regime-aware.

Mean-variance portfolio optimization via Conjugate Gradient — 40-60× faster than the legacy Neumann path (ADR-126 Phase 3, ADR-123 Wedge 8)

Optimize portfolio allocation using npx neural-trader mean-variance engine with risk constraints and rebalancing plan

Regulator-grade feature attribution for any LSTM/Transformer signal — single-entry PageRank ranks the top-K features that drove the prediction (ADR-126 Phase 6, ADR-123 single-entry PR)

Run a heavy neural-trader job (long walk-forward, big Monte-Carlo, parameter sweep, model training) on the Anthropic Managed Agent cloud runtime instead of locally

Run a historical backtest using npx neural-trader with Rust/NAPI engine (8-19x faster) and walk-forward validation; Ed25519-sign the result for paper→live tamper evidence (ADR-126 Phase 4)

Run 4-stem separation (vocals/drums/bass/other) on an existing production

Extract MIDI/score from an existing production

Run a mastering pass (LUFS loudness normalization + peak limiting) on an existing production

List the account's saved music productions with metadata and audio_url

Generate a new song from a creative brief (genre, mood, language, BPM, theme) via the cogmusic MCP create_production tool

One-time setup — mint a Cognitum Music personal access token and register the cogmusic MCP server with Claude Code

Validate pending migrations for foreign key consistency, rollback safety, and best practices

Create a new sequentially numbered database migration with up/down SQL files

ADR-152 — weighted similarity between two harness fingerprints (genome + score JSON). Returns overall score in [0,1] plus per-component breakdown (cosine over 9 numerics, categorical agreement over 4 enums, jaccard over agent_topology). Unblocks ADR-151 §3.2 Recommender, §3.3 Drift Detection, §3.5 Plugin Compat. Pure-TS, no `@metaharness/*` dep — preserves ADR-150's four architectural constraints.

Run `@metaharness/darwin security bench` (upstream "Darwin Shield" / ADR-155) — evolves a champion security-detection harness against a 10-vuln / 9-decoy corpus and grades it on TPR/FPR/patch-pass/repro/unsafe vs four baselines (B0 static, B1 LLM-single-pass, B2 fixed-agent, B3 Darwin-champion). Closest reference implementation for ruflo's own ADR-155 nightly self-learning security harness (PR

5-dimension harness readiness scorecard from `metaharness score {path}`. Returns harnessFit / compileConfidence / taskCoverage / toolSafety / memoryUsefulness + estCostPerRunUsd + scaffoldReady. Pure-read; subprocess invocation; degrades gracefully when MetaHarness is absent (ADR-150 architectural constraint).