Data analysis, machine learning, LLM apps, and prompt design (8.9K skills)
Prepare market history for pattern detection and comparison. Import opening, high, low, and closing prices plus volume, validate and normalize the values, then store them by symbol and date for later retrieval. It does not place orders.
Capture lessons from completed tasks, train patterns from explicit examples, and consolidate accumulated learning into long-term storage. Use SONA and MicroLoRA adaptation tools for individual domains, then verify results in the appropriate stores.
Inspect evolved prompt configurations, validate their structure, and view the system prompts they produce. You can also classify failures in run transcripts to support reviews before adopting or changing an agent policy.
Experiment with the policies that govern an AI agent, such as planning, memory, and tool use. Test one policy change per variant in a sandbox, an isolated environment, and promote only candidates with measured improvements.
Define reusable reasoning and decision patterns for agents, embed them in executable workflows, and review their performance. Share successful approaches, including conditional decision trees, stepwise problem solving, analogies, and evaluation from multiple perspectives.
Reuse patterns learned in another local project, or find, download, and publish patterns through an IPFS sharing registry. It supports confidence filters for local transfers and verification checks when downloading shared patterns.
Choose an agent and model tier using the task description and learned patterns. The skill can explain its recommendation, use compiler-based edits for supported simple changes, and record task outcomes to inform future routing decisions.
Create agents that learn from interactions and adapt their behavior over time. Monitor learning progress, review efficiency and accuracy metrics, trigger adjustments from feedback, and share acquired knowledge with other agents through dedicated tools.
Send proofs, planning questions, or difficult debugging tasks to DeepSeek's reasoning model. Receive the final answer and model-provided reasoning text in separate fields, with token usage reported for reviewing responses and costs.
Extract text, tables, and attributes from rendered web pages into structured data. Check extracted content for personal information and malicious instructions, and save reusable extraction templates for pages on the same host.
Assess tasks before execution to identify transformations eligible to skip LLM calls. Report the eligible share, recommended models for remaining tasks, and estimated upper costs, then save the routing results for later cost reports.
Check actual OpenRouter charges, API key limits, daily, weekly, and monthly spending, and account credits. Look up a single request by generation ID and report the billed model and provider returned by OpenRouter.
Review AI agent spending by conversation, with each session’s start time, message count, main model, and total cost. Use the results to identify expensive sessions and understand typical costs before planning longer work.
Estimate the cost per million tokens, units of model text, on hardware you own using power consumption, electricity rates, hardware costs, and measured generation speed. Compare utilization levels to find when local operation costs less than a hosted model.
Break down a Claude Code session's costs by assistant message to find expensive responses. Compare input, output, cache writes, cache reads, and models, with message cost percentiles providing context for investigating unusually costly activity.
Analyze the last seven days of agent and model token usage to identify unnecessary spending. Recommend changes to model selection, prompt caching, and overlapping work, with estimated dollar savings, percentage reductions, and implementation priorities.
Generate a spending report from recorded AI usage, with token counts and USD costs grouped by model, agent, and processing tier. Check budget utilization, remaining funds, and alerts for a selected period.
Retrieve compacted ReasoningBank context for cost-analysis queries and summarize retrieved memories, reported token savings, cache hit rate, and package availability. The savings figure is an estimate from the integration, not a measured comparison against a baseline.
Identify individual sessions with unusually high or low spending using the median and MAD, a measure of spread around it. Filter by time period, adjust detection thresholds, and optionally return a failure exit code for automated checks.
Compare actual AI session costs with estimates for using only Haiku, Sonnet, or Opus. The skill reports savings amounts and percentages, helping teams assess model selection costs while leaving differences in answer quality and retries outside the calculation.
Coordinates multiple AI agents with agentic-flow to run tasks in parallel or through dependent stages. It covers team communication structures, shared memory, workload balancing, failure recovery, and performance monitoring for complex workflows.
Build agents that record task outcomes and use past experience to choose future approaches. Implement pattern matching, strategy comparisons, persistent learning records, and metrics for checking how recommendations and results change over time.
Build advanced AgentDB workflows with database synchronization, multiple databases, and search that combines meaning with attribute filters. Covers similarity measures, diverse search results, and operational patterns for distributed AI applications.
Search document collections by meaning using embeddings, numerical representations of text. Compare text similarity, route queries among a few predefined patterns, and use a compressed search index when memory is limited.