Observability skills

This skill helps you accelerate RL-based LLM post-training with slime's Megatron-LM and SGLang for scalable data generation and rollout.

This skill helps you integrate PyTorch FSDP2 into training scripts with correct initialization, sharding, mixed precision, and DTensor-based checkpointing.

This skill ensures explicit user permission before any file deletion, preventing accidental loss by requiring approval for all delete operations.

This skill creates complete changelog announcements across detailed pages, summaries, and sidebar cards to publish release notes consistently.

This skill audits and updates the supported LLM model list in assets.py to align with litellm's registry, ensuring accuracy and zero duplicates.

This skill helps you build and customize fuzzers with LibAFL for advanced research, custom mutations, and fine-grained fuzzing control.

This skill helps identify timing side-channel leaks in cryptographic code by guiding statistical, dynamic, and formal analyses.

This skill audits GitHub Actions runtimes against thresholds, identifying near-timeout and timed-out jobs to prevent CI failures.

This skill enforces evidence-first debugging for coding tasks, guiding the agent through bounded escalation and dignified communication to resolve issues

This skill analyzes GitHub Actions artifact usage from JSON exports and flags oversized or stale artifacts to protect CI budgets.

This skill helps you manage and monitor Datadog resources by controlling monitors, dashboards, incidents, and more through Membrane.

This skill helps manage isolated sub-agent memory in OpenClaw by using scoped IDs, bounded results, and explicit cross-agent imports.

This skill enables interacting with Grafana dashboards, data sources, folders, and alerts via a managed-auth API gateway for monitoring and observability.

This skill analyzes meme coins for rug pull risk by evaluating contract safety, liquidity, distribution, and taxes to yield a 0-100 score.

This skill audits GitHub Actions trigger health by identifying highest failure-rate events across repos, workflows, and triggers to prioritize fixes.

This skill acts as a personal counselor by building a memory-informed profile, mirroring your patterns, and posing Socratic questions to foster self-awareness.

This skill analyzes GitHub Actions failures by day and hour to reveal recurring outage windows and staffing hotspots for actionables.

This skill provides real-time security monitoring and emergency controls for Sardis wallets, enabling circuit breakers and kill switches to prevent

This skill records and queries multi-channel activities with a 30-minute TTL, enabling smart summaries and cross-channel context for AI assistants.

This skill analyzes GitHub Actions runs to detect consecutive failure streaks by repo, workflow, and branch, surfacing hotspots for remediation.

This skill monitors crypto whale wallets across chains and alerts you to large transactions with customizable thresholds.