Subagent-Driven Development
HoangNguyen0403/agent-skills-standard
Runs a multi-task implementation plan by sending each task to a fresh implementer subagent, reviewing it independently, then reviewing the whole branch.
Enforce engineering rigor and verification loop for coding tasks using multi-agent debate and TDD.
Install Multi Agent Engineering Workflow
$ npx skills add nwjs/chromium.src --skill multi-agent-engineering-workflow -a claude-codeProject install by default; add -g for ~/.claude/skills/.
$ gh skill install nwjs/chromium.src multi-agent-engineering-workflow --agent claude-codeProject scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).
$ git clone --depth 1 https://github.com/nwjs/chromium.src.git skills-src && mkdir -p .claude/skills && cp -r skills-src/agents/skills/multi-agent-engineering-workflow .claude/skills/multi-agent-engineering-workflow && rm -rf skills-srcUse ~/.claude/skills/ instead of .claude/skills for a personal install. The folder must contain SKILL.md.
Claude Code skills documentation · loads skills from .claude/skills/
Install the "multi-agent-engineering-workflow" agent skill from https://github.com/nwjs/chromium.src/tree/main/agents/skills/multi-agent-engineering-workflow into .claude/skills/multi-agent-engineering-workflow/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "multi-agent-engineering-workflow", then confirm the skill loads.Claude Code copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$skill-installer install https://github.com/nwjs/chromium.src/tree/main/agents/skills/multi-agent-engineering-workflowType this inside Codex. $skill-installer <name> installs a curated skill from openai/skills. The installer writes to $CODEX_HOME/skills (default ~/.codex/skills). Restart Codex if the skill does not show up.
$ npx skills add nwjs/chromium.src --skill multi-agent-engineering-workflow -a codexProject install goes to .agents/skills/; add -g for ~/.codex/skills/.
$ gh skill install nwjs/chromium.src multi-agent-engineering-workflow --agent codexProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/nwjs/chromium.src.git skills-src && mkdir -p .agents/skills && cp -r skills-src/agents/skills/multi-agent-engineering-workflow .agents/skills/multi-agent-engineering-workflow && rm -rf skills-srcUse ~/.agents/skills/ instead of .agents/skills for a personal install.
Codex skills documentation · loads skills from .agents/skills/
Install the "multi-agent-engineering-workflow" agent skill from https://github.com/nwjs/chromium.src/tree/main/agents/skills/multi-agent-engineering-workflow into .agents/skills/multi-agent-engineering-workflow/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "multi-agent-engineering-workflow", then confirm the skill loads.Codex copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ npx skills add nwjs/chromium.src --skill multi-agent-engineering-workflow -a cursorProject install goes to .agents/skills/; add -g for ~/.cursor/skills/.
$ gh skill install nwjs/chromium.src multi-agent-engineering-workflow --agent cursorProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/nwjs/chromium.src.git skills-src && mkdir -p .cursor/skills && cp -r skills-src/agents/skills/multi-agent-engineering-workflow .cursor/skills/multi-agent-engineering-workflow && rm -rf skills-srcUse ~/.cursor/skills/ instead of .cursor/skills for a personal install.
Cursor skills documentation · loads skills from .cursor/skills/, .agents/skills/, .claude/skills/, .codex/skills/
Install the "multi-agent-engineering-workflow" agent skill from https://github.com/nwjs/chromium.src/tree/main/agents/skills/multi-agent-engineering-workflow into .cursor/skills/multi-agent-engineering-workflow/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "multi-agent-engineering-workflow", then confirm the skill loads.Cursor copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ gemini skills install https://github.com/nwjs/chromium.src.git --path agents/skills/multi-agent-engineering-workflow--scope user (default) or --scope workspace; --path is the subfolder of the repo that holds the skill; --consent skips the security confirmation prompt.
$ npx skills add nwjs/chromium.src --skill multi-agent-engineering-workflow -a gemini-cliProject install goes to .agents/skills/; add -g for ~/.gemini/skills/.
$ gh skill install nwjs/chromium.src multi-agent-engineering-workflow --agent gemini-cliProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/nwjs/chromium.src.git skills-src && mkdir -p .gemini/skills && cp -r skills-src/agents/skills/multi-agent-engineering-workflow .gemini/skills/multi-agent-engineering-workflow && rm -rf skills-srcUse ~/.gemini/skills/ instead of .gemini/skills for a personal install, then run /skills reload.
Gemini CLI skills documentation · loads skills from .gemini/skills/, .agents/skills/
Install the "multi-agent-engineering-workflow" agent skill from https://github.com/nwjs/chromium.src/tree/main/agents/skills/multi-agent-engineering-workflow into .gemini/skills/multi-agent-engineering-workflow/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "multi-agent-engineering-workflow", then confirm the skill loads.Gemini CLI copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ gh skill install nwjs/chromium.src multi-agent-engineering-workflowInstalls for Copilot at project scope by default; add --scope user for a personal install. Preview a skill first with gh skill preview. Needs GitHub CLI 2.90.0 or later (public preview).
$ npx skills add nwjs/chromium.src --skill multi-agent-engineering-workflow -a github-copilotProject install goes to .agents/skills/; add -g for ~/.copilot/skills/.
$ git clone --depth 1 https://github.com/nwjs/chromium.src.git skills-src && mkdir -p .github/skills && cp -r skills-src/agents/skills/multi-agent-engineering-workflow .github/skills/multi-agent-engineering-workflow && rm -rf skills-srcUse ~/.copilot/skills/ instead of .github/skills for a personal install. Commit .github/skills so cloud agent and code review can use it.
GitHub Copilot skills documentation · loads skills from .github/skills/, .claude/skills/, .agents/skills/
Install the "multi-agent-engineering-workflow" agent skill from https://github.com/nwjs/chromium.src/tree/main/agents/skills/multi-agent-engineering-workflow into .github/skills/multi-agent-engineering-workflow/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "multi-agent-engineering-workflow", then confirm the skill loads.GitHub Copilot copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ npx skills add nwjs/chromium.src --skill multi-agent-engineering-workflow -a opencodeOpenCode documents no install command of its own. Project install goes to .agents/skills/; add -g for ~/.config/opencode/skills/.
$ gh skill install nwjs/chromium.src multi-agent-engineering-workflow --agent opencodeProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/nwjs/chromium.src.git skills-src && mkdir -p .opencode/skills && cp -r skills-src/agents/skills/multi-agent-engineering-workflow .opencode/skills/multi-agent-engineering-workflow && rm -rf skills-srcUse ~/.config/opencode/skills/ instead of .opencode/skills for a personal install.
OpenCode skills documentation · loads skills from .opencode/skills/, .claude/skills/, .agents/skills/
Install the "multi-agent-engineering-workflow" agent skill from https://github.com/nwjs/chromium.src/tree/main/agents/skills/multi-agent-engineering-workflow into .opencode/skills/multi-agent-engineering-workflow/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "multi-agent-engineering-workflow", then confirm the skill loads.OpenCode copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
Facts
- Skill name
multi-agent-engineering-workflow- GitHub stars
- 160
- Token cost
- ~1.6k tokens
- SKILL.md length
- 636 words
- Files
- 102 (incl. references)
- Skills in repo
- 64
- Repo updated
- First seen
- Licence
- BSD-3-Clause
At a glance
Enforce engineering rigor and verification loop for coding tasks using multi-agent debate and TDD.
- Works in 7 steps: Tone Mandate (Signal-to-Noise) → Tool Agnostic Mandate → Environment Grounding Mandate → …
- Tasks that involve Multi-agent orchestration
- SKILL.md covers The Two-Path Model, Global Mandates & Invariants, Workflow Orchestration and Workspace Management & Isolation, plus 3 more sections
- Runs Python scripts from its folder
What it does
Multi Agent Engineering Workflow is an agent skill from nwjs/chromium.src. Enforce engineering rigor and verification loop for coding tasks using multi-agent debate and TDD.
Its SKILL.md is about 1.6k tokens, which your agent loads only when the skill is triggered. The skill folder holds 104 other files, including reference files (for example `EXAMPLES.md`, `PRESUBMIT.py` and `PRESUBMIT_test.py`).
It sits in Agent Workflows, covering Multi-agent orchestration and Test-driven development. The repository describes itself as: Chromium codebase with NW.js modifications. Based on https://chromium.googlesource.com/chromium/src.git. The licence is BSD-3-Clause.
When your agent uses it
- Tasks that involve Multi-agent orchestration
- Tasks that involve Test-driven development
Example prompts
- “/multi-agent-engineering-workflow”
Requirements
- Python 3
Workflow steps
7 steps, taken from the step headings in SKILL.md.
What it can do on your machine
Read from SKILL.md and the folder at commit a9e8946. It shows what the files ask for, not the result of running them.
Tool permissions
Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.
From allowed-tools in the SKILL.md frontmatter.
Runs code
Ships script files (Python, from the files we listed), which the agent can run.
From the folder's file list and the shell code blocks in SKILL.md.
Network
No URLs in SKILL.md.
From URLs in SKILL.md, links to its own repository left out.
Credentials
Names no API keys, tokens, secrets or passwords.
From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.
Context cost
Multi Agent Engineering Workflow loads about 1.6k tokens when it runs, and up to ~2.2k if it reads all its reference files. Until then it costs about 33 tokens; SKILL.md has 636 words of instructions outside code blocks.
Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.
Safety
Auto-check passed
The automated check found no risky patterns in SKILL.md.
Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); files beside SKILL.md are not scanned.
SKILL.md
The full file from nwjs/chromium.src at commit a9e8946, republished under its BSD-3-Clause licence (© nwjs). 636 words, ~1,592 tokens.
Download SKILL.mdSave it as
.claude/skills/multi-agent-engineering-workflow/SKILL.md (or your agent's skills folder). This skill also uses 101 other files; get the full folder from GitHub.- name
- multi-agent-engineering-workflow
- description
- Enforce engineering rigor and verification loop for coding tasks using multi-agent debate and TDD.
Multi-Agent Engineering Workflow
This skill acts as the high-level Orchestrator for the workflow protocol, a consensus-driven multi-agent framework designed to resolve complex software engineering problems. It coordinates scoping, TDD implementation, consensus reviews, and deployment by aggregating specialized sub-skills.
The Two-Path Model
The Orchestrator MUST select an execution path based on the task's complexity
and ambiguity (defined in project.workflow.json):
- FAST_PATH (Efficiency): Used for low-complexity, low-ambiguity tasks. Workflow: Scoping -> TDD/Direct Synthesis -> Single Auditor.
- RIGOR_PATH (Correctness): Default for high-complexity, high-ambiguity, or security-sensitive tasks. Workflow: Scoping -> TDD -> Consensus Review (Multiple Scanners).
Global Mandates & Invariants
All subagents invoked under the workflow protocol MUST adhere to these invariants to ensure harness compatibility and workspace safety.
1. Tone Mandate (Signal-to-Noise)
To eliminate conversational noise, conserve tokens, and maximize parsing stability, all agents (including the Orchestrator) MUST adopt a neutral, data-driven tone:
- Zero Preamble/Postamble: Sub-agents MUST NOT use conversational filler, greetings, or explanations of their work.
- Artifacts Only: If an agent's mandate is to generate JSON or code, its entire output MUST consist only of that raw data structure.
2. Tool Agnostic Mandate
The protocol instructions MUST remain tool-agnostic. Do not assume specific tool
names (e.g. update_topic, read_file, write_file). Use generic terms like
"read from disk," "save to disk," or "report status."
3. Environment Grounding Mandate
All sub-agents MUST read project.workflow.json#environment immediately upon
invocation to discover the active VCS (JJ or GIT) and Harness (JETSKI or
GENERIC_CLI). They MUST adjust their tool usage natively. All interim files
(drafts, reviews, logs) must be saved in the configured temp_directory to
prevent workspace pollution and minimize permission prompts.
Workflow Orchestration
Stage 0: Initialization & Scoping
- Ground the environment, discover active VCS, and verify tool availability.
- Investigate the initial request and write
project.workflow.jsonto configure the project (VCS, temp directories, target files) and define the goal. - Read the
temp_directoryandexecution_pathfrom the generatedproject.workflow.json. - Clean up any leftover state files (
tdd_state.workflow.json,review_state.workflow.json,constraints.workflow.json) in the configuredtemp_directoryto ensure a clean start.
Stage 1: TDD Implementation
- Invoke the
multi-agent-tdd-implementation
skill (passing any active
constraints.workflow.jsonif iterating). - Wait for completion and verify that the synthesis build/test target compiles.
Show full SKILL.md (283 more words)Show less
Stage 2: Consensus Review & Audit
- If FAST_PATH: Invoke multi-agent-code-review but select only a single auditor.
- If RIGOR_PATH: Invoke multi-agent-code-review selecting the "Big Three" scanners (Security, Performance, Auditor) and any domain specialists.
- Read the consolidated
verdictandnext_stagefromreview_state.workflow.json:- ACCEPT (or
next_stage: COMPLETED): Transition to Stage 3. - REJECT (or
next_stage: SYNTHESIS): Ifoscillation_detected == falseand global iterations < 3, loop back to Stage 1. Else, escalate to the user. - ESCALATION (
next_stage: ESCALATION): Pause and present the conflict report to the user.
- ACCEPT (or
Stage 3: Deployment & Cleanup
- Invoke the multi-agent-release-manager skill to format code, run presubmits, and upload the final CLs.
Workspace Management & Isolation
- Interim File Isolation: Place all draft files (
*.workflow,*.workflow.*) in the configuredtemp_directory(e.g.agents/skills/multi-agent-engineering-workflow/.temp/). - Cleanup: The release skill MUST delete the temporary directory at the end of a successful run.
- VCS & Staging Workflows: Upgrades to workflow configuration files (via multi-agent-skill-trainer) must be branched and uploaded as separate secondary CLs.
Reference Guides
- Routing and Specialization: Consult ROUTING.md to understand how tasks are routed to specialized sub-agents based on file patterns and complexity.
- JSON Configuration Contract: Consult EXAMPLES.md for the
exact schema and examples of the configuration JSON files
(
project.workflow.json,review_state.workflow.json, etc.). - Testing Protocol: Consult SKILL_TEST_PLAN.md and SKILL_TEST.md for verification procedures and unit tests.
- Harness & Orchestration Patterns: Consult orchestration_patterns.md to understand how the Orchestrator adapts to centralized (Jetski) or decentralized (MAS CLI) environments.
Roadmap & Architecture TODOs
- TODO(Expanded Engineering Phases):
- Integrate upstream design phases (e.g., generating design documents, class diagrams, and sequence diagrams).
- Integrate downstream verification phases (e.g., automated code-coverage
gating via
experimental-code-coverage-config-validatorand fuzzing viafuzzing).
Evaluation & Testing
When modifying this skill's workflow, routing, or schemas, ensure that the corresponding Promptfoo evaluation test suite is updated and passing:
© nwjs, BSD-3-Clause. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file
Files
SKILL.md and 101 other files (references) in agents/skills/multi-agent-engineering-workflow of nwjs/chromium.src.
- SKILL.md
- .gitignore
- .ruff.toml
- .style.mdformat
- EXAMPLES.md
- OWNERS
- PRESUBMIT.py
- PRESUBMIT_test.py
- README.md
- ROUTING.md
- SKILL_TEST.md
- SKILL_TEST_PLAN.md
- check_json_format.py
- personas/ai/llm.json
- personas/ai/mas.json
- personas/auxiliary/build.json
- personas/auxiliary/concurrency.json
- personas/auxiliary/i18n.json
- … and 84 more
Open the folder on GitHubat commit a9e8946
Compare with similar skills
Multi Agent Engineering Workflow next to the 5 skills that share the most tags, products or categories with it. Stars are the repository's; “used in” counts other GitHub owners with a copy.
| Skill | Stars | Used in | Tokens | Auto-check | Licence | Repo updated |
|---|---|---|---|---|---|---|
| Multi Agent Engineering Workflow this skillnwjs/chromium.src | 160 | — | ~1.6k | Automated safety check: Pass | BSD-3-Clause | |
| Subagent-Driven DevelopmentHoangNguyen0403/agent-skills-standard | 571 | — | ~1.3k | Automated safety check: Pass | MIT | |
| SPARC Development Methodologyruvnet/ruflo | 74k | 2 repos | ~829 | Automated safety check: Pass | MIT | |
| TDDglebis/claude-skills | 390 | — | ~8.4k | Automated safety check: Pass | MIT | |
| Deep Planpiercelamb/deep-plan | 101 | — | ~4.8k | Automated safety check: Pass | MIT | |
| Tbdjlevy/strif | 131 | — | ~3.5k | Automated safety check: Pass | MIT |
Similar skills
SPARC Development Methodology
ruvnet/ruflo
Applies the SPARC method (specification, pseudocode, architecture, refinement, completion) with 17 specialized modes and multi-agent orchestration, from research to deployment.
TDD
glebis/claude-skills
This skill should be used when the user wants to implement features or fix bugs using test-driven development.
Deep Plan
piercelamb/deep-plan
Creates detailed, sectionized, TDD-oriented implementation plans through research, stakeholder interviews, and multi-LLM review.
Tbd
jlevy/strif
Git-native issue tracking (beads), coding guidelines, knowledge injection, and spec-driven planning for AI agents.
- Official
Cao Provider
awslabs/cli-agent-orchestrator
Create a new CLI agent provider for CAO (CLI Agent Orchestrator).
More from nwjs/chromium.src
All 64 skills in this repoAnalyzing SQL Traces
nwjs/chromium.src
Extracts raw trace data from Perfetto traces, runs arbitrary SQL queries for custom follow-up analysis, and applies expert cognitive principles (Tiered Flow Analysis, Semantic Mismatch, Redundancy)…
Chrome Performance Optimizer
nwjs/chromium.src
Autonomous multi-agent performance optimization loop for Chromium and V8.
Automated Tracing
nwjs/chromium.src
Automated Tracing & Performance Telemetry in Chromium using Perfetto and Telemetry benchmarks.
Chrome Releases
nwjs/chromium.src
Queries Chrome commit, version, release, and milestone metadata.
Chromium Docs
nwjs/chromium.src
Search and reference Chromium documentation from the local docs index, including design docs, APIs, and development guides.
Gn Deps Debugging
nwjs/chromium.src
Diagnose Chromium GN dependency and include-visibility failures, including BUILD.gn deps/publicdeps, DEPS include rules, private headers, and circular dependencies.
Related
Categories
Questions about Multi Agent Engineering Workflow
What does Multi Agent Engineering Workflow do?
Enforce engineering rigor and verification loop for coding tasks using multi-agent debate and TDD. src. Enforce engineering rigor and verification loop for coding tasks using multi-agent debate and TDD.
When should I use Multi Agent Engineering Workflow?
Multi Agent Engineering Workflow fits situations like: tasks that involve Multi-agent orchestration; tasks that involve Test-driven development.
How do I install Multi Agent Engineering Workflow in Claude Code?
Run `npx skills add nwjs/chromium.src --skill multi-agent-engineering-workflow -a claude-code`. Or copy the skill folder (agents/skills/multi-agent-engineering-workflow in nwjs/chromium.src) into .claude/skills/multi-agent-engineering-workflow in your project. Claude Code loads it when a task matches its description.
How do I install Multi Agent Engineering Workflow in Codex?
Run `npx skills add nwjs/chromium.src --skill multi-agent-engineering-workflow -a codex`. Or copy the skill folder (agents/skills/multi-agent-engineering-workflow in nwjs/chromium.src) into .agents/skills/multi-agent-engineering-workflow in your project. Codex loads it when a task matches its description.
Can I use Multi Agent Engineering Workflow in Cursor, Gemini CLI or GitHub Copilot?
Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add nwjs/chromium.src --skill multi-agent-engineering-workflow -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/multi-agent-engineering-workflow, .gemini/skills/multi-agent-engineering-workflow, .github/skills/multi-agent-engineering-workflow and .opencode/skills/multi-agent-engineering-workflow in your project.
What does Multi Agent Engineering Workflow need to run?
Going by SKILL.md and its folder, Multi Agent Engineering Workflow needs Python for the scripts in its folder. Our summary lists: Python 3.
Does Multi Agent Engineering Workflow access the network?
SKILL.md contains no URLs. Any network use would come from the scripts or tools the agent runs. This is read from the text; nothing was executed.
Is Multi Agent Engineering Workflow safe to install?
Our automated static check of SKILL.md found no risky patterns, such as piping downloads into a shell, reading credential files or hidden Unicode. It is not a guarantee. Review the folder before installing.
What licence does Multi Agent Engineering Workflow use?
Multi Agent Engineering Workflow is published under the BSD-3-Clause licence (the repository's licence). It allows redistribution, so the full SKILL.md is shown on this page.
How many tokens does Multi Agent Engineering Workflow use?
About 1.6k tokens (SKILL.md is roughly 6.4k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full. Its references folder adds about 586 tokens, read only when the agent opens those files.
What are the alternatives to Multi Agent Engineering Workflow?
Skills that share tags, products or a category with Multi Agent Engineering Workflow: Subagent-Driven Development (HoangNguyen0403/agent-skills-standard, 571 stars), SPARC Development Methodology (ruvnet/ruflo, 74k stars), TDD (glebis/claude-skills, 390 stars) and Deep Plan (piercelamb/deep-plan, 101 stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.
Who maintains Multi Agent Engineering Workflow?
nwjs (a GitHub organization) maintains it in nwjs/chromium.src, which has 160 GitHub stars. The repository holds 64 skills in this directory. The repository was last updated on October 3, 2026.
Source: nwjs/chromium.src on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.