skills/wave/SKILL.md
Expert Wave multi-agent pipeline orchestrator development including manifest configuration, pipeline authoring, persona management, and CLI operations
npx skillsauth add re-cinq/wave waveInstall this skill globally with one command. Works with Claude Code, Cursor, and Windsurf.
3 of 9 scanners reported clean
Some scanners were skipped, did not run, or reported a non-clean status. Review each row below.
Go CLI orchestrating multi-step AI workflows. Each pipeline step runs in an isolated workspace under a named persona with enforced permissions, artifact chaining, and contract validation.
wave.yaml)apiVersion: v1
kind: WaveManifest
metadata:
name: my-project
adapters:
claude:
binary: claude
mode: headless
output_format: json
project_files: [CLAUDE.md, .claude/settings.json]
default_permissions:
allowed_tools: [Read, Write, Edit, Bash]
personas:
navigator:
adapter: claude
system_prompt_file: .agents/personas/navigator.md
model: sonnet # opus | sonnet | haiku
permissions:
allowed_tools: [Read, Glob, Grep, "Bash(git log*)", "Bash(git status*)"]
deny: ["Write(*)", "Edit(*)", "Bash(git commit*)", "Bash(git push*)"]
runtime:
workspace_root: .agents/workspaces
max_concurrent_workers: 5
default_timeout_minutes: 30
relay:
token_threshold_percent: 80
strategy: summarize_to_checkpoint
summarizer_persona: summarizer
skill_mounts:
- path: .agents/skills/
.agents/pipelines/<name>.yaml)kind: WavePipeline
metadata:
name: gh-pr-review
release: true
steps:
- id: analyze
persona: navigator
dependencies: []
memory:
strategy: fresh
inject_artifacts:
- step: prior-step
artifact: artifact-name
as: local-filename # available at artifacts/<as>
exec:
type: prompt
source: |
Analyze the code for: {{ input }}
output_artifacts:
- name: analysis
path: output/analysis.json
type: json
handover:
contract:
type: json_schema # or: command
schema_path: .agents/contracts/analysis.schema.json
source: output/analysis.json
must_pass: true
on_failure: retry
max_retries: 2
| Persona | Role |
|---------|------|
| navigator | Read-only codebase exploration |
| craftsman | Code implementation & testing |
| implementer | Full execution specialist |
| philosopher | Architecture & specification |
| auditor | Security review & QA |
| summarizer | Context compaction for relay |
| reviewer | Quality review & validation |
| github-commenter | Post GitHub issue comments |
Permission syntax: Read, Write(.agents/specs/*), Bash(git log*), Bash(*) in deny
output_artifacts: # produced by this step
- name: analysis
path: output/data.json
type: json
memory:
inject_artifacts: # consumed from prior step
- step: analyze
artifact: analysis
as: analysis_data # available at artifacts/analysis_data
wave run gh-pr-review "Review auth module"
wave run hotfix --dry-run
wave run <pipeline> --from-step <step> --force # resume
wave do "fix the login bug" # ad-hoc task
wave meta "implement user auth" # dynamic pipeline
wave status --all
wave logs --step investigate --errors --follow
wave artifacts --step analyze --export ./out
wave list pipelines
wave list runs --limit 20 --run-status completed
wave validate --pipeline gh-pr-review
wave clean --all --keep-last 5 --dry-run
wave cancel --force
{{ input }}, {{ timestamp }}dependencies run concurrentlymax_retries)auto (TUI on TTY), json (NDJSON), text, quietgo test ./...
go test -race ./...
go run ./cmd/wave run hello-world "test"
wave run hello-world "test" --mock
For exhaustive patterns, examples, and advanced usage see:
references/full-reference.md
testing
Test generation, coverage analysis, and test strategy
development
Terraform infrastructure as code — providers, modules, state management
tools
Tailwind CSS — utility-first styling, responsive design, component patterns
testing
Security scanning, vulnerability assessment, and hardening