Total 56,830 skills, AI & Machine Learning has 9448 skills
Showing 12 of 9448 skills
Decision guide for delegating to hui-style subagents. Tells the main thread WHEN to spawn `huicrew-investigator` (locate code), `huicrew-builder` (1-2 file edit), or `huicrew-reviewer` (diff review) instead of doing the work inline or using vanilla `Explore`. Subagent output is hui-compressed so the tool-result injected back into main context is ~60% smaller — main context lasts longer across long sessions. Trigger: "delegate to subagent", "use huicrew", "spawn investigator/builder/reviewer", "save context", "compressed agent output".
Use when the user wants to build, initialize, validate, optimize, or refactor a model-powered assistant, internal tool, automation, evaluator, or workflow from a business scenario or common problem statement, including project-structure refactors or starter skeletons that may separate model setup, prompt config, and orchestration, even if the request also mentions a UI, app shell, or local model service such as Ollama, and it is still unclear whether the solution should stay a single request, add supporting capabilities, or become orchestration. The user does not need to mention Agently explicitly.
AI SDLC Conventional Commit workflow. Use when an AI assistant drafts, validates, reviews, or fixes commit messages in this repository, especially when commits must include SDD spec references, validation summaries, or safe conventional commit subjects. Supports `--quick-flow` for fast assumption-driven execution and `--full-flow` for question-driven verified execution.
AI SDLC resumable task-runtime workflow. Use when an AI assistant needs to start or resume a versioned delivery run, select dependency-ready work, enforce step, failure, and token budgets, retry safely, persist exact stop reasons, recover state from an append-only journal, or require commit evidence at task boundaries. Supports `--quick-flow` for deterministic local runs and `--full-flow` for strict transition review.
AI SDLC declarative workflow planning. Use when an AI assistant needs to validate a versioned workflow, plan typed dependency steps, evaluate bounded conditions, enforce approval gates, attach deterministic hooks, detect cycles, or create safe dependency waves with sequential fallback when host concurrency or isolation is unavailable. Supports `--quick-flow` and `--full-flow`.
AI SDLC repository delivery-graph and evidence-freshness workflow. Use when an AI assistant needs to index lifecycle traceability, resolve end-to-end paths, report gaps or orphans, register evidence identity, propagate stale dependencies, or calculate fresh evidence coverage. Supports `--quick-flow` for deterministic local analysis and `--full-flow` for strict trace and evidence review.
Build NPC AI in Unreal Engine 5 with Behavior Trees and Blackboards: composites (Selector/Sequence), tasks, decorators, services, and running the tree from an AIController. Use when creating enemy/NPC AI, BT_/BB_ assets, custom BTTask or BTService nodes, or when the user mentions Behavior Tree, Blackboard, AIController, BTTask, decorator, or service.
Deep-dive diagnosis of how your AI agent behaves in production. Explores LangWatch analytics and traces end to end to map failure patterns, dissatisfied users, token cost hotspots, edge cases, behavior changes, and outliers, then delivers an HTML report where every finding links to real example traces. Use when you want to truly understand what your agent is doing in production.
Create, modify, run, inspect, analyze, and report Python experiments that use liblaf.cherries. Use when Codex needs to work under exp/YYYY/mm/dd/group-name/, write or edit numbered scripts in src/, run them with CHERRIES_NAME and CHERRIES_TAGS, inspect Cherries/Comet logs and generated assets, or write Markdown reports in docs/.
Interactive exploration of ideas through Socratic Q&A. Produces progressive documents that serve as lightweight pre-PRDs feeding into research.
Execute a single DAG step as an autonomous background sub-agent. Sibling of phase-running for DAG plans produced by v-planning. Reads a step-<n>.md file directly, atomically claims it via frontmatter status, runs the three-bucket Success Criteria, and reports back. Spawned by v-implementing or by /run-step.
Improve (or bootstrap) an AGENTS.md / CLAUDE.md file using `<important if>` conditional blocks so the agent actually attends to the right guidance at the right time. Use this skill whenever the user mentions AGENTS.md, CLAUDE.md, agent instructions, project rules for AI, "my claude config", onboarding docs for agents, or asks to tighten / shorten / audit / rewrite an existing one — even if they don't explicitly say the filename. Also use when the user complains that an agent keeps ignoring their project rules.