l3io-arch-review
Engineering-standards architecture guardrails and review.
Skill by ravensorb
Chalk Labs
Need help deploying l3io-arch-review?
Chalk Labs sets up agents like this inside real teams: connected to your tools, tested on your data, with a human in the loop where it matters.
Install
Works with Claude Code, Claude apps, Codex, Cursor, Gemini CLI, GitHub Copilot, VS Code, OpenCode, Goose, Windsurf. Read the repository before you give any tool access to your data.
npx agentmarket install skill/ravensorb-l3io-arch-review
Asks which of your tools to add it to (Claude Code, Codex, Cursor, VS Code, Windsurf, Gemini CLI, Claude Desktop) and shows the change first. See https://agentmart.dev/cli.
Get the weekly diff
The weekly diff: one email a week with new agent tools, notable updates and grade changes, plus security alerts for tools you follow. Double opt-in, one-click unsubscribe.
About
Use cases: documents, design, devops
Reviews
Write a reviewNo reviews yet. Installed it? Reviews come from verified installers only. How reviews work
Permissions
What this skill can do and what it asks for, from its SKILL.md. Checked Oct 25, 2026.
- Tools
- No tools found
- Secrets
- None asked for
- Install scripts
- None found
- Runs as
- Python package
Network hosts it talks to 1
codeload.github.com· found in install command
Files it touches
~/.claude/skills/l3io-arch-review
What's inside
- Skill name
l3io-arch-review- Description
- Engineering-standards architecture guardrails and review. Use when the user wants to apply best-practice engineering standards — separation of concerns, reuse, design-by-contract, testability, dependency/GA policy, unified correlated logging, documentation with diagrams, plus per-stack rules (Python uv/poetry, Node LTS, .NET self-contained, GitHub Actions, Docker digest-pinning/SBOM, PowerShell 7 major-scoped, shell strict-mode) — at new-project design time, during an architectural review, or when recording a technology/architecture decision.
- License
- MIT
- Instructions
- 98 lines in SKILL.md · view on GitHub
Before installing
- Ships executable scripts
Files (23)
- SKILL.md
- adr-template.md · assets/
- agent-instructions.md · assets/
- customize-architect.md · assets/
- module-help.csv · assets/
- module-setup.md · assets/
- module.yaml · assets/
- customize.toml
- module.yaml
- payload-manifest.json
- config-resolution.md · references/
- review-report.md · references/
- standards-core.md · references/
- standards-docker.md · references/
- standards-dotnet.md · references/
- standards-github-actions.md · references/
- standards-nodejs.md · references/
- standards-powershell.md · references/
- standards-python.md · references/
- standards-shell.md · references/
- merge-config.py · scripts/
- merge-help-csv.py · scripts/
- write-module-config.py · scripts/
SKILL.md (first lines)
# Architecture Standards & Review
You apply LiquidLogicLabs engineering standards to a project's architecture. You operate in
three modes — **design guardrails** (new project), **review** (audit an existing design or
system), and **decision support** (weigh options, record an ADR). The standards are the
single source of truth in `references/standards-*.md`; you do not invent rules, you apply and
cite them.
A clean review on non-trivial scope is suspect — walk every principle before you conclude.
## Conventions
- Bare paths (e.g. `references/standards-core.md`) resolve from the skill root.
- `{skill-root}` resolves to this skill's installed directory (where `customize.toml` lives).
- `{project-root}`-prefixed paths resolve from the project working directory.
- Findings use severity **BLOCKER / MAJOR / MINOR**, and each names: principle, location, remediation.
## On Activation
Resolve config through BMad core's resolver — full contract in
`references/config-resolution.md`:
```bash
uv run --python 3.11 {project-root}/_bmad/scripts/resolve_config.py --project-root {project-root}
```
If the resolver is missing or fails, BMad core is not installed here — stop and tell the
user to run the BMad installer. Read `core.communication_language` (default `English`) and
present all output in that language; take module settings from `modules.l3io-arch` and the
artifact paths from `modules.l3io-pm` (see the contract for defaults).
An absent `modules.l3io-arch` section is normal and is **not** a first-run trigger — this
module declares no required settings. Load `assets/module-setup.md` only when the user
explicitly passes `setup` or `configure`.
1. **Always load** `references/standards-core.md` (the universal charter).
2. **Detect the stack(s)** in scope and load the matching overlay(s):
- Python (`pyproject.toml`, `uv.lock`, `poetry.lock`) → `references/standards-python.md`
- Node.js (`package.json`) → `references/standards-nodejs.md`
- .NET/C# (`*.csproj`, `*.sln`) → `references/standards-dotnet.md`
- GitHub Actions (`.github/workflows/*`) → `references/standards-github-actions.md`
- Docker / PowerShell / shell → load the matching overlay if present.
When the user names a stack explicitly, honor that over auto-detection.
3. **Pick the mode** from the user's intent (or ask if ambiguous). The three words
`design`, `review` and `decision` name the modes directly — they are the `action`
values of this module's three `assets/module-help.csv` rows, so that is what bmad-help
hands a user who picks one from the menu. An argument that is not one of them is still
read as intent, never rejected:
### Mode A — Design guardrails (new project)
Walk `standards-core.md` §1–10 plus each loaded overlay as a **design checklist**. For the
target project, produce:
- A boundaries/architecture sketch honoring separation of concerns (§1) with at least a C4
context + one flow diagram (Mermaid preferred, ASCII fallback — §10).
- The initial **ADR set** for every load-bearing call (stack choice, key dependencies, any
preview/non-GA use, logging stack, deployment mode) using `assets/adr-template.md` —
numbered and placed as in Mode C.
- A `/docs` skeleton across the architectural / developer / operational axes (§10).
- A dependency policy note (GA-over-beta §8, maintenance+license §7).
### Mode B — Architectural review
Audit the target (whole design, a component, or a diff). For **each** principle in the loaded
files, emit findings (`references/review-report.md` shapes the output). A finding is only valid
when it states **severity · principle · location · concrete remediation**. Roll findings up
into a report with an executive summary and a severity-graded table. Gate: BLOCKER and MAJOR
findings must be resolved or ADR-justified; MINOR auto-defers to the backlog.
### Mode C — Decision support
Given a decision in play, identify the principle(s) in tension, weigh the options against them,
recommend, and **record an ADR** (`assets/adr-template.md`) at
`{project-root}/docs/adr/NNNN-slug.md`, the one ADR home. When
`{project-root}/_bmad/scripts/pm-status/__main__.py` exists, take the number from the l3io-pm
register:
`uv run {project-root}/_bmad/scripts/pm-status/__main__.py adr-reserve --state-root {implementation_artifacts}/state --epic n/a --slug {slug} --adr-dir {project-root}/docs/adr`
prints it. Without l3io-pm there is no register, so use the highest number in `docs/adr/`
plus one. That is safe only for a single writer, which is the only case without the
register, and `adr-reserve` skips those numbers later. Fill `Epic:` (`n/a` outside an epic)
and `Departs from spec:` (a `path#anchor`, or `n/a`). Never let a load-bearing call go
unrecorded.
## Wiring into core BMad (customization)
This module is also designed to be wired **into** the core `bmad-architecture` and
`bmad-code-review` skills so the standards apply automatically. `assets/customize-architect.md`
documents the `bmad-customize` overlay to author in a consuming project. Offer to set it up.
## Output
- **Review** → report to `{implementation_artifacts}` (or the path the orchestrator passes).
- **Design/Decision** → ADRs under `{project-root}/docs/adr/` (numbered by `adr-reserve` when l3io-pm is installed — Mode C) and the docs skeleton.
End every non-interactive run with:
`DONE — Blocker: N, Major: N, Minor: N | BLOCKED: [reason] | FAILED: [reason]`From ravensorb/bmad-extensions, MIT.
More from ravensorb/bmad-extensions
- 0.570score
l3io-execute
ravensorb
Run the l3io-pm plan — full, single epic, or single sprint.
SkillAFreshness: AClaude Code +9 - 0.570score
l3io-help
ravensorb
Read project state and recommend the exact next l3io action, or list every installed l3io skill.
SkillAFreshness: AClaude Code +9 - 0.570score
l3io-plan
ravensorb
Validate readiness, elaborate stories, estimate, build dependency graph, and produce an executable plan.
SkillAFreshness: AClaude Code +9 - 0.570score
l3io-sec-redteam
ravensorb
Red team security analysis agent.
SkillAFreshness: AClaude Code +9 - 0.570score
l3io-setup
ravensorb
Sets up the LiquidLogicLabs PM module in a project.
SkillAFreshness: AClaude Code +9 - 0.570score
l3io-sync
ravensorb
Bidirectional sync between l3io-pm state and GitHub Issues.
SkillAFreshness: AClaude Code +9 - 0.570score
l3io-doctor
ravensorb
Migration and housekeeping utilities for BMad artifacts and l3io-pm state.
SkillAFreshness: AClaude Code +9 - 0.556score
l3io-pm-plan
ravensorb
DEPRECATED forwarder for /l3io-plan — renamed in 3.1.3.
SkillAFreshness: AClaude Code +9 - 0.556score
l3io-pm-help
ravensorb
DEPRECATED forwarder for /l3io-help — renamed in 3.1.3.
SkillAFreshness: AClaude Code +9 - 0.556score
l3io-pm-sync
ravensorb
DEPRECATED forwarder for /l3io-sync — renamed in 3.1.3.
SkillAFreshness: AClaude Code +9 - 0.556score
l3io-pm-execute
ravensorb
DEPRECATED forwarder for /l3io-execute — renamed in 3.1.3.
SkillAFreshness: AClaude Code +9
Signals
- Stars
- 0
- Forks
- 0
- Open issues
- 0
- Last commit (this folder)
- Oct 22, 2026 (8 days ago)
From GitHub (fetched Oct 24, 2026).
Details
- License
- MIT
- Repository
- github.com/ravensorb/bmad-extensions
- First seen
- Oct 24, 2026
- Last checked
- Oct 24, 2026
- Source
- GitHubgithub-graphql
More from ravensorb
- 0.570score
l3io-execute
ravensorb
Run the l3io-pm plan — full, single epic, or single sprint.
SkillAFreshness: AClaude Code +9 - 0.570score
l3io-help
ravensorb
Read project state and recommend the exact next l3io action, or list every installed l3io skill.
SkillAFreshness: AClaude Code +9 - 0.570score
l3io-plan
ravensorb
Validate readiness, elaborate stories, estimate, build dependency graph, and produce an executable plan.
SkillAFreshness: AClaude Code +9 - 0.570score
l3io-sec-redteam
ravensorb
Red team security analysis agent.
SkillAFreshness: AClaude Code +9 - 0.570score
l3io-setup
ravensorb
Sets up the LiquidLogicLabs PM module in a project.
SkillAFreshness: AClaude Code +9 - 0.570score
l3io-sync
ravensorb
Bidirectional sync between l3io-pm state and GitHub Issues.
SkillAFreshness: AClaude Code +9
Done for you
Rather not set it up yourself?
Chalk Labs scopes, builds and runs agents for your team: the task audit, the eval harness, the integrations and the weekly ops report. Fixed scope, priced before we start.
See how we build AI agents →Read the case studies
Chalk Labs is a two-founder AI and growth studio. We build AI agents, CRMs and MVPs for Web3 and AI companies, and have shipped 50+ product and token launches.
- launches shipped
- 50+
- launches shipped
- media placements
- 500+
- media placements
- founders on every account
- 2
- founders on every account
Press we've placed: Forbes · CoinDesk · Decrypt · Cointelegraph · The Block
Rahil Jain, Co-founder, Chalk Labs. Ex-Goldman Sachs engineer. You talk to the person who builds it.
Get l3io-arch-review running in your business
Tell us what you want to automate. A founder replies within one working day with a plan, not a pitch deck.
Cite this page
l3io-arch-review skill. Agent Mart. https://agentmart.dev/skills/ravensorb-l3io-arch-review (last updated 2026-10-24)