Autoresearch Coding
Run a metric-driven coding experiment loop. Use when improving a repo through repeated hypotheses, benchmark-guided changes, validation, keep/discard decisions, or safe workflow enhancement. Requires a goal, metric command, direction, editable paths, protected paths, and budget.
Tutien
Run a private, persistent xianxia coding-reflection mode with evidence-backed classification and an open-ended, repository-specific living chronicle. Use only when the user invokes /tutien; once enabled, carry its cultivation-novel voice across every reply until /tutien off or an explicit request ends the mode.
Threat Model Security Review
Run a dependency-free, threat-model-led application security review of a repository, scoped component, code diff, or supplied vulnerability claim. Use when auditing source code, authentication, authorization, input handling, filesystem or network access, sensitive data paths, trust boundaries, or security regressions. Produces evidence-backed attack paths, explicit validation status, coverage gap…
Wait What
The last message did not land - stop and re-pitch it in plain language, in the user's language, until it lands.
The Creator
Create original but workable art, designs, interfaces, methods, processes, systems, or other solutions through ten cumulative creativity levels. Use only when the user explicitly invokes the-creator, names a creator level from 1 to 10, or directly asks to use the ten-level Creator framework for logical and functionally acceptable invention.
Agentshield Security Review
repository-native security review workflow for Claude Code, Cursor, Codex, OpenCode, Grok, and Kimi projects using the AgentShield concept. Use when auditing or hardening agent configuration, skills, prompts, hooks, MCP servers, permissions, commands, plugin files, AGENTS.md, CLAUDE.md, or repository-level AI coding-agent surfaces. Prefer this skill before merging changes to .claude, .cursor, .co…
Clearthought
Structured reasoning for complex coding work. Use when requirements are ambiguous, a design/debugging decision has multiple valid paths, or a large request needs to be split into safe implementable tasks.
Clone Website
Plan, build, and verify an authorized website clone or local research replica from a public URL, screenshots, owner export, or existing repository. Use when cloning a website, reproducing page structure or visual design, migrating an owned site, choosing a suitable frontend and backend level, or measuring UI and interaction parity. Requires explicit rights handling, bounded capture, neutralized p…
Prompt Sharpener
Sharpen a rough prompt into a precise one - same intent, sharper not longer - then execute the sharpened version immediately in the same turn.
Mermaid
Generate styled Mermaid diagrams (31 types) from requirements - flowcharts, sequence, class, ER, state, Gantt, kanban, timeline, XY charts, architecture, and more. Also use when writing documentation (offer to include diagrams), when debugging or explaining a workflow (offer a visual flow map with risky zones highlighted red), or whenever the user asks for a chart, diagram, or visualization. Adap…
Clean Delivery
Deliver one observable behavior slice through Specify, Code, Clean, Architect, Harden, and Verify gates with proportional rigor and reproducible proof. Use when the user asks for extreme craftsmanship, clean-code delivery, TDD, architecture boundaries, or a high-confidence implementation workflow.
Visual Design Loop
Run an iterative visual design improvement loop for UI/product polish tasks. Use when a request involves screenshots, rendering, visual review, frontend polish, layout/typography/color refinement, design QA, or a Claude/Codex loop goal that touches a visible product surface. Score the visual gate first and wait for user approval before starting loops.
Daily Workflow Curator
Propose daily improvements to project AI rules, skills, workflows, prompts, AgentShield posture, and backbone.yml. Use when the user asks to auto-enhance, update workflow, improve rules, or keep the vibe coding kit current. Propose diffs only; do not silently write.
Proofline Orchestration
Govern complex agent work with an original four-role contract, independent challenge, explicit escalation signals, evidence-bound handbacks, and optional provider adapters. Use when the user asks for Proofline, role-separated orchestration, an empowered dissenting reviewer, multi-agent governance, or a Paseo-compatible coordination setup. Keep simple work sequential and never claim independence t…
Parallel Analysis
Fan out 2-5 independent read-only analysis lanes across the repo using provider-ready native subagents or configured adapters, then merge the lane reports and verify them with a refutation pass. Use for repo-wide questions, large uncommitted-diff reviews, multi-doc reading, impact analysis, or consistency audits. Before dispatch, resolve the project's Default, Auto, or Custom orchestration prefer…
Vibekit Init
Initialize or repair Minimal Vibe Coding Kit in an existing project. Use for first prompt, backbone.yml setup, CLAUDE.md merge, AGENTS.md merge, Cursor rules setup, Codex skill setup, OpenCode setup, Grok setup, or Kimi setup. Always show requirements and a diff before writing.
Path Sensitive Shell Safety
Guardrail workflow for shell changes that use path variables, remote base directories, repo folders, checkout repair, or destructive filesystem commands such as rm, mv, cp, rsync, find -delete, git clean, or git reset.
Memento
Leave yourself a note before closing a multi-day task, then read it back at the start of the next session - it's the only memory that survives the gap.
Claim
Vet a request to bring something new into the repo - validate every URL and reference against official sources, check fit with existing rules, conventions, and skills, confirm anything unclear with the user, then integrate it and document how to use it.
Reviewing 4p Priorities
P0-P4 issue and task triage for vibe coding. Use when reviewing bugs, risks, feature requests, agent-surface changes, or deciding which work must happen first.
Graph Engineering Verified Orchestration
Design and run bounded dependency graphs for complex engineering work, with explicit artifacts, isolation, objective verification, resource budgets, retries, rollback, merge ownership, and human gates. Use when a task has at least three bounded work items, two or more genuinely independent branches, and plausible time/cost benefit or coordination-risk reduction; also use when the user asks for gr…
Sequential Thinking
Evidence-led step-by-step reasoning for complex coding tasks. Use to decompose work, revise assumptions, compare alternatives, verify hypotheses, and choose the next safe action without exposing private chain-of-thought.
Coding Level
Set the coding-explanation register. /coding-level N (0=ELI5, 1=Junior, 2=Mid, 3=Senior, 4=Tech Lead, 5=God); active until reinvoked. A project default may be recorded in backbone.yml conventions.custom_rules.