Cloud, CI/CD, monitoring, and deployment (6.8K skills)
Investigate missing replies, incomplete output files, and stalled callbacks in CCB by tracing requests through execution and delivery. Use the recorded evidence to choose a retry, fresh submission, or acknowledgement repair and verify the resulting state.
Configure CCB agents, providers, workspaces, and window layouts through its project configuration file. Validate edits, preview reload effects, and identify affected agents, with reload execution available when the user explicitly requests applying the change.
Prepare, publish, and audit CCB releases across GitHub and npm. Align version files, README content, release notes, downloadable assets, and automated checks, then verify that the published package and repository homepage reflect the intended release.
Investigate CCB problems such as stalled agents or missing replies using runtime diagnostics, message records, logs, and tmux panes, the agents’ terminal views. Produce a concise diagnosis with evidence, confidence, and a suggested next action.
Consult a historical reference describing how agent capacity responsibilities are divided. It outlines evidence for requested profiles, counts, blockers, and release readiness, while assigning actual capacity management and work routing to the supervisor or runner script.
Recover CCB agents after runtime, provider, or API problems. Check active and pending work before clearing context, validating and reloading configuration, or restarting one affected agent at a time, then verify its runtime state.
Investigate a named CCB agent that appears stuck or disconnected using runtime records, task history, and terminal evidence. Apply supported, limited repairs, verify recovery, and prepare a redacted incident record for review before any GitHub issue submission.
Set up FireRed-OpenStoryline on a local computer, from installing Python dependencies and downloading models and assets to configuring API access, starting services, and checking that the installation works.
Investigate failed automated checks on a pull request or branch, identify actionable errors from logs or linked services, and apply small fixes. Recheck results after each push and report changes, current status, and next steps.
Push Python package changes to GitHub with version updates chosen from pending commits. Keep version fields aligned and trigger an existing CI workflow, an automated checking and publishing process, to release the package on PyPI.
Investigate Nexu production issues through the Datadog Logs API. Search crash events, error output, startup activity, and API failures, then summarize findings by time, running instance, and deployed version to help trace related events.
Publish static websites to nexu.space using Wrangler and Cloudflare Pages. Supports initial launches and later updates, with cache revalidation settings that help prevent outdated pages and assets from appearing without changing source files.
Diagnose failing GitHub Actions and Codecov checks on Comet pull requests using the latest commit, failed job logs, and merge status. It distinguishes code failures from environment issues or outdated checks and reports evidence with the smallest necessary action.
Prepare Comet releases by checking actual versions and branch changes, writing user-facing changelog entries, coordinating Chinese and English website documentation, and running checks appropriate to the changes. Report release contents, verification results, and remaining uncertainties.
Prepare a new ralph-orchestrator release by updating workspace and dependency versions, building and testing the project, and pushing the commit and version tag. Then track the automated release workflow that creates binaries and publishes packages.
Use Cloudflare’s official command-line tool to manage DNS and storage resources and develop or deploy Workers. The skill covers finding commands, checking request details, previewing changes, authentication, and choosing between cf and Wrangler.
Investigate CPU usage and memory allocation in deployed Cloudflare Workers and Durable Objects. The skill analyzes captured performance profiles, identifies costly functions where evidence allows, and proposes small code changes with ways to check their effect.
Plan, configure, and troubleshoot Cloudflare One deployments for secure application access and traffic filtering. Review identity, devices, routes, and logs, then develop changes with validation steps, phased rollout, and rollback plans.
Plan migrations from existing VPN and web security platforms to Cloudflare One. Organize exported settings and logs, map access rules to target resources, identify unsupported behavior, and prepare pilot testing, rollout, and rollback plans.
Run and troubleshoot Wrangler commands in existing configured projects. Support local development, preview environments, deployments, secrets, and resource management while checking the installed version, target environment, and relevant validation steps.
Set up and troubleshoot Cloudflare K2, a durable log that stores events between producers and consumers. The skill guides stream creation, sending records from Workers or HTTP, retention settings, and receiving records through subscriptions.
Choose Cloudflare products for applications, AI agents, storage, networking, and security. This skill connects project requirements to suitable services, explains how they fit together, and identifies product-specific skills or documentation for implementation.
Deploy applications and manage infrastructure with Cloudflare, Docker, Google Cloud, and Kubernetes. Get guidance on choosing services, packaging applications into containers, automating deployment workflows, troubleshooting, and reviewing access controls and security settings.
Prepare remediation steps for Azure Advisor recommendations using the affected resource’s details. Guidance comes from the Advisor remediation tool or public Microsoft Learn documentation, with warnings, verification steps, and clear source attribution.