By nanana291
Universal route, ownership, handoff, and delivery-gate scaffold that complements Superpowers.
npx claudepluginhub nanana291/agent-system --plugin agent-systemEmit the brain dedupe contract and list merge candidates
Emit the delivery gate contract from the executable enforcement path
Emit a DELIVERY GATE block and decide whether the work is blocked or ready
Emit a HANDOFF block for the current owner stage
Emit a TASK LOCK using the active profile and selected route
Generate Luau release notes from a baseline comparison. Links feature changes, brain entries, risk score trends, and delivery proof into a readable changelog.
Generate a visual comparison matrix between two Luau script versions (baseline vs modified). Shows feature parity, remote changes, risk deltas, and local pressure.
Detect features present in an original Luau script that are missing or degraded in a modified version. Reports lost features, degraded features, and new additions.
Summarize data flow in a Luau script: how locals feed functions, how functions feed remote calls, and how UI wiring triggers logic. Produces a flow graph in text form.
Block or approve Luau script delivery based on quality thresholds. Checks risk score, remote safety, loop safety, character lifecycle, and feature parity before allowing delivery.
Map all remote calls in a Luau script: names, arguments, pcall status, fire method, frequency, and handler coverage. Reports missing handlers and unsafe remote patterns.
Generate a Luau script scaffold with safety patterns pre-wired. Supports new script, new feature, LibSixtyTen section, and V2 migration templates.
Compare the current metrics snapshot against a prior reference
Compare the current metrics snapshot against recent history
Show or capture the workspace metrics trail
Show the active profile, task types, classifications, and artifact contract
Classify the task, pick the active profile route, and list required skills and artifacts
Train multiple agents, inspect recent history, and sync training memory
Emit the upgrade apply contract and create or refresh the upgrade session snapshot
Learn per-agent upgrade lessons and persist the upgrade snapshot
Compare a historical upgrade session against the current target
Print the latest upgrade learning state and per-agent lesson summary
Print the latest upgrade learning state without writing files
Reapply the stored upgrade state without re-inference
Run the learning-aware upgrade pipeline for the active profile
Review routing structure, ownership boundaries, profile shape, and host compatibility for architectural drift.
Migrate Luau scripts between UI frameworks (Obsidian → LibSixtyTen), versions (V1 → V2), or structural patterns. Handles cloning, renaming, feature parity, and regression proof.
Analyze obfuscated, decompiled, or encoded Luau scripts. Reconstructs logic, identifies anti-tamper, detects backdoors, and maps real functionality behind obfuscation.
Review completed work for regression risk, missing proof, and delivery-gate compliance.
Classify a task, select the profile route, and return the ordered agent stages with required skills and artifacts.
Verify that the required routing artifacts exist and match the current task stage before allowing delivery.
Automatically suggest and apply consistent tags to brain entries based on content analysis. Ensures searchable, well-organized brain data.
Validate the delivery gate, required proof, and blocker state before work is marked complete.
Assign each active domain to one owner, define what cannot change next, and emit a handoff that preserves control boundaries.
Check and enforce cross-executor compatibility for Luau scripts. Detects executor-specific APIs, suggests fallbacks, and generates compatibility reports.
Search and apply reusable Luau patterns before writing or modifying code. Prevents re-discovering known solutions.
Pre-modification safety checklist for any Luau file. Verifies backups, brain entries, remote safety, UI init, and loop safety before touching code.
Define how to verify Luau scripts without executing them — source-level checks, invariant assertions, pattern validation, and structural tests.
Choose the active profile from the repository context, host context, and explicit user request.
Comprehensive UI/UX design plugin for mobile (iOS, Android, React Native) and web applications with design systems, accessibility, and modern patterns
Uses power tools
Uses Bash, Write, or Edit tools
Qiushi Skill: methodology skills for AI agents guided by seeking truth from facts, with Claude Code, Cursor, OpenClaw, Codex, OpenCode, and Hermes guidance.
Standalone image generation plugin using Nano Banana MCP server. Generates and edits images, icons, diagrams, patterns, and visual assets via Gemini image models. No Gemini CLI dependency required.