architecture-engineer
Analyze BACKS architecture, detect structural drift, evaluate proposed changes against the rules file, and produce architecture decision records.
Skill by Tcuzzo
Chalk Labs
Need help deploying architecture-engineer?
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/tcuzzo-architecture-engineer-ef890b
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: design, security
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
Network hosts it talks to 1
codeload.github.com· found in install command
Files it touches
~/.claude/skills/architecture-engineer
What's inside
- Skill name
architecture-engineer- Description
- Analyze BACKS architecture, detect structural drift, evaluate proposed changes against the rules file, and produce architecture decision records. Use before approving any boundary-changing change — new modules, dependency direction shifts, or moves that merge concerns the project keeps separate. Trigger words: architecture, structural drift, dependency direction, ADR, boundary change, responsibility leak, architecture review, design drift, rules compliance.
- License
- MIT
- Instructions
- 67 lines in SKILL.md · view on GitHub
Files (1)
- SKILL.md
SKILL.md (first lines)
# Architecture Engineer — drift detection + ADR authorship The BACKS AIOS mirror of the proven BACKS skill at `skills/architecture_engineer/SKILL.md`. Use before approving any change that touches a module boundary, dependency direction, or shared seam. ## When to run - A new module is proposed — does it deserve to exist, or does the work belong in an existing primitive? - An existing primitive is being moved, renamed, or absorbed — what breaks? - A new dependency direction is being added — does it cross a known forbidden boundary? - An ADR is needed — a durable decision warrants a record. ## The review recipe (six steps) 1. **Fixed point.** Identify the request, the diff or revision under review, and the repository's current architecture + contribution rules. Trace the affected entrypoints, dependencies, data flow, configuration, persistence, and callers before judging the change. 2. **Declared vs runtime.** Treat declared architecture (docs, ADRs) and runtime wiring (the actual seam) as separate evidence. Report when they disagree — that's the drift. 3. **Local rule check.** Run the project rule check first (`backend/services/authority_surface_classifier.py` for BACKS; equivalent for other repos). A change that registers as authority_surface + requires_approval=true is a gate — surface the approval row to the operator, never bypass it. 4. **Structural review.** Check the change for: - responsibility and dependency direction across boundaries - duplicated orchestration, import cycles, new abstractions with no caller - configuration embedded in code when the project owns a config seam - registrations/routes/adapters without a reachable runtime implementation - merging concerns the project deliberately keeps separate - drift from explicit project rules (permission/authority boundaries) 5. **Severity + smallest fix.** For each finding: severity, exact evidence location, consequence, smallest architecture-level remedy. Do not invent numeric targets from another repo. 6. **ADR when durable.** Write an ADR only when a durable decision needs recording. Use the project's existing ADR location and format. Capture: status, context, decision, considered alternatives, consequences, superseded decisions. Never manufacture an ADR for a reversible implementation detail. ## What never to do - **Do not enforce stale numeric targets or conventions from another repo.** Cite the local rule or concrete failure mode behind every violation. Label unproven design concerns as risks, not defects. - **Do not bypass the authority surface classifier.** If it returns requires_approval=true, the change belongs in a proposal row, not a direct commit. The classifier's fail-closed default is the law (inv 2). - **Do not refactor adjacent code during a review.** Stay on the diff under review; surface adjacent issues but do not touch them. ## Cross-references - The BACKS source of truth: `skills/architecture_engineer/SKILL.md` and `skills/architecture-engineer/SKILL.md` (hyphenated) in the BACKS repo. - The authority classifier: `backend/services/authority_surface_classifier.py`. - The fix-shit companion: this skill + `skills/fix-shit` for understanding- first repair. - Companion skills: `wayfinder`, `intent-compiler`, `seam-engineering`, `turn-and-burn` — they are the other half of the operator's named stack.
From Tcuzzo/HydraAgent_public, MIT.
More from Tcuzzo/HydraAgent_public
- 0.598score
backs-pack-sync
Tcuzzo
Use when the BACKS AIOS plugin pack has drifted from the BACKS source skills/ — a new BACKS skill exists that the IDE cannot discover because the plugin pack…
SkillAFreshness: A★ 2Claude Code +9 - 0.598score
conscientious-objector
Tcuzzo
Use when an IDE agent (Cursor, Codex, Claude Code) is operating inside BACKS / BACKSAIOS and a human-quality watchdog is needed to verify the harness is being…
SkillAFreshness: A★ 2Claude Code +9 - 0.598score
co-health
Tcuzzo
Use when the Conscientious Objector (or any watchdog) needs a health check on the operator's own cadence — proven live, on demand, never silently.
SkillAFreshness: A★ 2Claude Code +9 - 0.598score
clean-code-gauntlet
Tcuzzo
Use when hardening or landing any build (an agent, a service, a library) and you want a deterministic quality bar instead of a line-by-line review.
SkillAFreshness: A★ 2Claude Code +9 - 0.598score
cam-iot-inventory
Tcuzzo
Use when inventorying cameras/IoT/Wi-Fi on the live BACKS Secure console — MAC-sticky camera identity, honest wifi-scan, LO5 inventory≠picture.
SkillAFreshness: A★ 2Claude Code +9 - 0.598score
blind-tribunal
Tcuzzo
Use when an autonomous change needs an independent grade before landing and no human is in the loop.
SkillAFreshness: A★ 2Claude Code +9 - 0.598score
blind-eval
Tcuzzo
Use before landing anything where taste or output quality is the question and a test cannot decide it.
SkillAFreshness: A★ 2Claude Code +9 - 0.598score
bounded-loops
Tcuzzo
Use before starting any loop that can retry, poll, iterate, or call an external API — agent loops, repair loops, schedulers, watchers.
SkillAFreshness: A★ 2Claude Code +9 - 0.598score
architecture-engineer
Tcuzzo
Evaluate structure against the project that actually exists, not a remembered diagram or a generic architecture preference.
SkillAFreshness: A★ 2Claude Code +9 - 0.598score
absorb
Tcuzzo
Use when you need a capability an open-source project already provides — adopt it and re-engineer it as a native skill instead of inventing a duplicate.
SkillAFreshness: A★ 2Claude Code +9 - 0.593score
bughunt
Tcuzzo
Use when running a bounded parallel bug hunt that maps seams, verifies findings adversarially, and closes whole flaw classes.
SkillAFreshness: A★ 2Claude Code +9 - 0.593score
agent-build
Tcuzzo
Use when building an agent, bot, worker, or autonomous service with deterministic primitives, typed IO, and bounded loops.
SkillAFreshness: A★ 2Claude Code +9
Signals
- Stars
- 2
- Forks
- 0
- Open issues
- 0
- Last commit (this folder)
- Oct 6, 2026 (25 days ago)
From GitHub (fetched Oct 24, 2026).
Details
- License
- MIT
- Repository
- github.com/Tcuzzo/HydraAgent_public
- Website
- tcuzzo.github.io/
- First seen
- Oct 24, 2026
- Last checked
- Oct 24, 2026
- Source
- GitHubgithub-graphql
More from Tcuzzo
- 0.598score
bounded-loops
Tcuzzo
Use before starting any loop that can retry, poll, iterate, or call an external API — agent loops, repair loops, schedulers, watchers.
SkillAFreshness: A★ 2Claude Code +9 - 0.598score
architecture-engineer
Tcuzzo
Evaluate structure against the project that actually exists, not a remembered diagram or a generic architecture preference.
SkillAFreshness: A★ 2Claude Code +9 - 0.598score
absorb
Tcuzzo
Use when you need a capability an open-source project already provides — adopt it and re-engineer it as a native skill instead of inventing a duplicate.
SkillAFreshness: A★ 2Claude Code +9 - 0.598score
backs-pack-sync
Tcuzzo
Use when the BACKS AIOS plugin pack has drifted from the BACKS source skills/ — a new BACKS skill exists that the IDE cannot discover because the plugin pack…
SkillAFreshness: A★ 2Claude Code +9 - 0.598score
blind-eval
Tcuzzo
Use before landing anything where taste or output quality is the question and a test cannot decide it.
SkillAFreshness: A★ 2Claude Code +9 - 0.598score
blind-tribunal
Tcuzzo
Use when an autonomous change needs an independent grade before landing and no human is in the loop.
SkillAFreshness: A★ 2Claude 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 architecture-engineer 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
architecture-engineer skill. Agent Mart. https://agentmart.dev/skills/tcuzzo-architecture-engineer-ef890b (last updated 2026-10-24)