practice-cycle
Move a non-trivial task through practice, observation, reflection, and the next smallest test without forcing a generic plan or feedback artifact. Use when a…
Skills
Find an Agent Skill for your task. Read the guide and copy the install command.
Updated 11-11 00:16
Agents & models
26,093 results
project-cognition
Build, verify, or selectively refresh a source-linked working model of a substantial project when later agents or tasks need to reuse its architecture,…
skill-engineering
Design, rewrite, review, and test agent skills that improve a repeated agent action or judgment. Use when creating a skill, rewriting a SKILL.md, improving a…
skill-incubator
Incubate or improve an Agent skill from theory, existing methods, and practice; design the core methodology and an executable workflow; and revise candidates and the incubation method through layered evaluation, isolated experiments, and feedback. Use for skill creation where the method still needs development or its effectiveness still needs verification; packaging an established method, ordinary instruction editing, full skill architecture, and domain task execution each have their corresponding methods.
light-figure-planning
Based on the paper's content, plan which figures and tables should be made, where they should be inserted, and what role each plays. Use when the user needs paper figure and table planning. Figures and tables are not limited to statistical charts; they also include real effect figures of datasets, model output examples, case demonstrations, interpretability visualizations, etc. Plan framework diagrams, technical roadmap diagrams, dataset schematic diagrams, model architecture diagrams, algorithm flowcharts, result comparison/ablation/sensitivity figures, real effect figures, statistical tables/comparison tables, etc., and judge by reviewer standards which are must-do, …
light-orchestrator
Orchestrator. When the user gives a large task spanning multiple stages (such as "from this dataset to a paper" or "help me take this project from research to submission") or asks to "continue / it just broke / take over / restore context", plan or restore a staged pipeline, invoke the corresponding skills stage by stage along the CONVENTIONS stage mainline, and set mandatory checkpoints between stages (decision points +…
light-result-analysis
Conduct a detailed, professional, and in-depth analysis of the executed results, experimental data, model outputs, and chart results. Use this when the user's experiment has finished running, needs the data interpreted, or asks "what do these results mean." Do not merely describe whether they are good or bad; instead explain why, which results prove the method is effective, which expose problems, which anomalies need investigation, which patterns can become paper highlights, and which require additional experiments.
skill-template
[One-paragraph description of what this skill does. Include when it should be triggered and what problem it solves. Be specific — a good description helps…
excalidraw-skill
Programmatic canvas toolkit for creating, editing, and refining Excalidraw diagrams via MCP tools with real-time canvas sync. Use when an agent needs to (1)…
survey-generation
Generate complete academic survey papers using multi-LLM parallel outline generation, RAG-based subsection writing, citation validation, and local coherence…
gh-issues
Fetch GitHub issues, spawn sub-agents to implement fixes and open PRs, then monitor and address PR review comments. Usage: /gh-issues [owner/repo] [--label…
oracle
Best practices for using the oracle CLI (prompt + file bundling, engines, sessions, and file attachment patterns).
voice-call
Start voice calls via the OpenClaw voice-call plugin.
ccx-flow
Enable or disable CCX Flow, allowing an external long-running process to inject context into the current CC main session.
engage
Autonomous execution mode triggered by the word "engage". Use when the user has finished planning and wants the agent to execute autonomously without further…
ferris-search-tools
CRITICAL: Use for ferris-search MCP tool usage. Triggers on: web_search, fetch_web_content, fetch_github_readme, fetch_csdn_article, fetch_juejin_article,…
flowtrace
Manage or review business records in FlowTrace through its MCP server, including projects, deliveries, requirements, stages, bugs, personal work, action items,…
thinker-model-pack-validation
Use when the user provides a quantized ONNX model exported by Linger and requests packaging with tpacker, tuning Thinker memory and splitting parameters to reduce hardware operator splitting, verifying inference consistency between Linger and Thinker via tvalidator, or generating a minimal reproduction report.
ls
liustack skill stack entry point and routing. Hand the current problem to it, and it selects the appropriate ls-* skill based on context and executes according to its process. Entry point and router for the liustack skill stack. Hand it any problem and it routes to the right ls-*…
ls-systems-thinking
Systems thinking: draw feedback loops (reinforcing/balancing), find the bottleneck and leverage point, rather than treating the symptom instead of the root cause. Systems thinking. Map reinforcing and balancing feedback loops, find the constraint and the leverage point. Use when the…
sync-upstream-gateway
Align the Go gateway implementation in this repository with the upstream LobeHub reference implementations (device-gateway / agent-gateway) and the application-side client protocol. Invoke when you need to sync the upstream protocol, catch up with the latest implementation, or align to a specified LobeHub tag/commit.
earnings-analysis
Create professional equity research earnings update reports (8-12 pages, 3,000-5,000 words) analyzing quarterly results for companies already under coverage.…
initiating-coverage
Create institutional-quality equity research initiation reports through a 5-task workflow. Tasks must be executed individually with verified prerequisites -…
lbo-model
This skill should be used when completing LBO (Leveraged Buyout) model templates in Excel for private equity transactions, deal materials, or investment…