2.5K skills · sorted by relevance
Inspect CPU, memory, disk, and accelerator information alongside container and scheduler limits. Produce a JSON snapshot with identifying details removed and conservative worker and memory estimates for planning local workloads, without running stress tests.
Answer questions about past decisions, usual practices, and changes over time by reasoning across Hindsight memories. When information spans multiple memory banks, combine the answers and identify their sources, while acknowledging gaps or contradictions.
Save facts, decisions, preferences, and task outcomes to Hindsight long-term memory. The skill turns what was learned into short notes with context and reasons, choosing either a Bot’s own memory store or a shared store.
Retrieve relevant Hindsight memories before planning or answering a request. Searches the agent’s own and shared memory stores, plus other relevant stores when needed, to bring past decisions, user preferences, and project context into the task.
Carry context between scheduled or event-triggered routine runs using Hindsight memory. Add recall and recording steps to routine instructions so each run can use previous findings, resume unfinished work, compare changes, and leave guidance for the next run.
Design agent memory that preserves facts, preferences, and conversation history across sessions. Plan storage, retrieval, and retention separately for each organization or user, with checks for unauthorized access, outdated information, conflicting records, and retrieval quality.
Guides users through removing a stored fact with Codewhale’s Context Lens memory controls. It identifies the intended fact, explains the forget action, and distinguishes confirmed deletion from an action that has not yet completed.
Choose SGLang Diffusion settings for a specific model, GPU, and memory budget. Compare generation speed, memory use, and output quality to select command-line options, including CPU offloading, multiple GPUs, caching, and reduced precision.
Investigate process hollowing, where malicious code replaces a legitimate process in memory. Compare memory with disk files, examine process relationships, and correlate network activity to document suspicious processes and guide containment.
Investigate signs of Mimikatz credential theft using command lines, access to Windows authentication memory, file indicators, and known modules in memory. Correlate endpoint and security logs to assess suspicious alerts, document evidence, and recommend response actions.
Write and improve C code for systems and embedded projects, with explicit memory ownership, pointer handling, and error checks. The skill also covers build files, unit tests, debugging, and performance measurement.
Compress large language models to 4-bit weights to reduce GPU memory use and accelerate inference. Covers loading compressed models, configuring quantization, and using Transformers or vLLM, with guidance on hardware requirements and backend choices.
Fine-tune large language models with LoRA and QLoRA while limiting GPU memory use. Guides training settings and adapter management, using small sets of task-specific weights that can be saved, switched, or merged into a base model.
Build collaborating AI teams with CrewAI by defining roles, tasks, and shared context. Configure ordered execution or manager-led delegation, add memory and tools, and create workflows that branch based on events or results.
Ask questions about your journal, Obsidian vault, or other text notes, and continue conversations that remember earlier chats. The companion uses cognee memory with local databases, reads your notes without changing them, and saves conversations for later reference.
Compare assigned work from connected trackers with TASKS.md, review stale tasks, and fill gaps in stored context about people and projects. An optional comprehensive scan finds possible missed commitments in recent activity, with changes confirmed by the user.
Add GBrain memory to an existing agent or connect a hosted brain while preserving current instructions and identity. Start with storage and keyword search that need no API key, then verify saving, retrieval, and recall across conversations.
Finish a NanoClaw v1-to-v2 migration after the migration script completes. The skill checks real message routing, assigns owner access, coordinates legacy memory migration, reviews container settings, and helps decide how to carry over customizations.
Save agent memory and outputs in Cua Volume, shared storage that persists beyond a Space, the working environment. Hand files to the user, read shared references, and check synchronization, version history, and conflicting edits.
Shows recent work sessions for the current project as a timeline, newest first. It includes session identifiers, start times, status, recorded observation counts, and available highlights or decisions, using only history returned by the memory tool.
Resume work using the latest recorded session matching the current project directory. The skill surfaces unanswered questions first, summarizes relevant decisions and files, and provides a concrete next step while checking that records belong to the selected project.
Investigate slow Node.js, Python, and Go applications by measuring CPU use, memory, database queries, and data transfer. Identify bottlenecks, inspect frontend bundle sizes, run load tests, and compare measurements before and after changes to verify improvements.
Reconcile project documentation, agent rules, authorized memory, and workspace leftovers with current code and runtime behavior. Prepare a clear handoff by correcting stale information, identifying authoritative explanations, and reporting what is verified, pending, or awaiting cleanup approval.
Investigate crashes, out-of-memory errors, and device initialization failures on GKE TPU v6e nodes. The skill checks logs and custom metrics collection for conflicting operations, then guides temporary mitigation by disabling an identified collector.