Skills¶
Skills teach a coding agent (Claude Code, Codex, Gemini CLI, Cursor,
Antigravity and others) the graph-agents-cli lifecycle. This page is generated from each
skill's SKILL.md at build time.
graph-agents-cli setup # install the CLI and the skills
graph-agents-cli update # refresh the skills, upgrade the CLI
setup installs the skills of the release your CLI's version names, and falls back to the
copy bundled in the wheel when it cannot fetch them.
Installation & setup covers the options (a workspace
install, another source, a specific agent). Once installed, ask your coding agent to "use
graph-agents-cli to build ..." and the graph-agents-cli-workflow skill leads.
graph-agents-cli-workflow¶
Entrypoint for building LangGraph agents with graph-agents-cli. Always active: provides the full workflow (understand, scaffold, build, evaluate, deploy, observe), process deference to a project's declared process, the spec-before-code gate, code preservation rules, the never-change-the-model rule, human approval before deploy, and the 3-strikes loop breaker.
Picked up by requests like “develop an agent”, “build an agent with LangGraph”, “build a LangGraph agent”, “run the agent locally”, “debug agent code”, “test an agent”, “evaluate an agent”, “deploy an agent to Kubernetes”, “monitor an agent”.
Source: SKILL.md · References: brainstorming, commands, extension, internals, spec-template, terminology
Skill version 0.3.1
graph-agents-cli-deploy¶
Covers the deployment modes (direct local-load, direct registry, helm-push, argocd), the dev/staging/prod environments, the secrets procedure and rotation, the Argo CD PR flow and the single production merge gate, the required GitHub environment and branch-protection settings, infra check, local-load dev clusters, and the disconnected profile.
Picked up by requests like “deploy an agent”, “deploy to Kubernetes”, “deploy to our cluster”, “set up CI/CD”, “set up Argo CD”, “configure secrets”, “rotate a key”, “promote to production”, “check cluster prerequisites”, “deploy to kind/k3s/minikube”, “deploy air-gapped”, “troubleshoot a deployment”.
Not for agent code (graph-agents-cli-langgraph-code), evaluation (graph-agents-cli-eval), scaffolding (graph-agents-cli-scaffold), or tracing (graph-agents-cli-observability).
Source: SKILL.md · References: github-settings, gitops, kubernetes, secrets
Skill version 0.3.1
graph-agents-cli-eval¶
Covers the enforceable eval gate (one rule, case statuses, exit codes), the dataset schema, deterministic expect checks, judge and quality metrics, the local-versus-disconnected distinction, and eval submit. Applies to any graph-agents-cli project.
Picked up by requests like “run an evaluation”, “evaluate my agent”, “write an eval dataset”, “add an eval case”, “analyze eval failures”, “compare eval results”, “set a quality threshold”, “why did eval exit 1”, “upload evals to LangSmith”.
Not for agent code (graph-agents-cli-langgraph-code), deployment (graph-agents-cli-deploy), or scaffolding (graph-agents-cli-scaffold).
Source: SKILL.md · References: dataset_schema, metrics-guide
Skill version 0.3.1
graph-agents-cli-langgraph-code¶
Covers create_agent and StateGraph, tools with the API_CALLS declaration, memory vs postgres checkpointers and thread_id, streaming, interrupts, subgraphs, init_chat_model provider switching, the fake provider for tests, the auth policy adapter, the API client and api-policy.yaml, and telemetry opt-in.
Picked up by requests like “write agent code”, “build an agent with LangGraph”, “add a tool”, “add a node to the graph”, “use a checkpointer”, “stream events”, “add human-in-the-loop”, “add a subgraph”, “switch the model provider”, “implement the auth policy”, “call an external API from a tool”.
Not for scaffolding (graph-agents-cli-scaffold), evaluation (graph-agents-cli-eval), or deployment (graph-agents-cli-deploy).
Source: SKILL.md · References: langchain-models, langgraph, template-contract
Skill version 0.3.1
graph-agents-cli-observability¶
Covers the TRACING_ENABLED opt-in, the TRACE_CAPTURE metadata versus full policy, LangSmith versus OTLP, the hashed principal id, run records, and what is never captured by default.
Picked up by requests like “set up tracing”, “enable LangSmith”, “send traces to our collector”, “monitor my agent”, “debug production traffic”, “see what the agent sent to the model”, “log prompts”.
Not for deployment (graph-agents-cli-deploy) or agent code (graph-agents-cli-langgraph-code).
Source: SKILL.md · References: langsmith, otel
Skill version 0.3.1
graph-agents-cli-scaffold¶
Covers graph-agents-cli create, scaffold enhance, and scaffold upgrade with every flag, the valid runtime x checkpointer x target combinations, prototype semantics, the registry default, the authentic-baseline rule for upgrade, and the files upgrade never touches.
Picked up by requests like “create an agent project”, “start a new LangGraph project”, “build me a new agent”, “scaffold a project”, “add Kubernetes deployment”, “add CI/CD to my project”, “add Argo CD”, “enhance my project”, “upgrade my project”.
Not for writing agent code (graph-agents-cli-langgraph-code) or deployment operations (graph-agents-cli-deploy).
Source: SKILL.md · References: flags
Skill version 0.3.1