architecture-methodology-4ec7ec — independently scanned and version-tracked by SaferSkills.
SaferSkills independently audited architecture-methodology-4ec7ec (Agent Skill) and scored it 100/100 (green). The audit ran 55 deterministic rules across Security, Supply Chain, Maintenance, Transparency, and Community; it found 0 high-severity and 0 lower-severity findings. The full rule-by-rule trace and per-finding evidence are below. Free, methodology-open.
Findings & checks · 0 flagged
Every scanned point with the score it earned and what moved between them.
First recorded scan — no prior version to compare against.
The primary manifest — the file an agent reads to learn what this artifact does.
Any conversation about system architecture, technology choices, component design, PRD analysis, or solution design.
| Phase | Name | Purpose |
|---|---|---|
| 1 | Evaluation | Analyse PRDs, extract functional and non-functional requirements, identify constraints and risks. No technology or component decisions yet. |
| 2A | Pattern | Select the architecture pattern (e.g. event-driven, microservices, monolith) and justify the choice. |
| 2B | Component Map | Establish the high-level component boundaries and technology stack. |
| 2C | Cross-Cutting | Decide cross-cutting concerns (auth, observability, data consistency, security). These become constraints for all Phase 3 component designs. |
| 3 | Components | Design individual components sequentially. Each component must be explicitly accepted before the next begins. |
| 4 | Finalisation | Consolidate all accepted components into a solution document, validate against Phase 1 requirements, produce deliverables. |
.arch/state.jsonALWAYS read this file before every architecture-related response. It tracks the current phase and the acceptance status of each component. NEVER rely on conversation memory for phase state.
Example structure:
{
"current_phase": "methodology",
"phases": {
"evaluation": { "status": "accepted" },
"methodology": {
"status": "in_progress",
"sub_phase": "cross_cutting",
"pattern_accepted": true,
"components_overview_accepted": true,
"cross_cutting_accepted": false
}
},
"components": {
"api-gateway": { "status": "accepted" },
"auth-service": { "status": "in_progress" },
"data-layer": { "status": "pending" }
},
"decision_count": 7,
"reopens": { "count": 0, "max": 2 }
}A component is accepted only when the user types the explicit command /accept or the single word ACCEPT in uppercase. Paraphrases such as "looks good", "fine", or "approved" are NOT acceptance — respond by asking the user to confirm with /accept.
.arch/state.jsonBefore advancing to the next phase, verify all criteria are met:
Phase 1 → Phase 2:
phases.evaluation.status set to "accepted" in statePhase 2 → Phase 3:
phases.methodology.pattern_accepted: true)phases.methodology.components_overview_accepted: true)phases.methodology.cross_cutting_accepted: true)phases.methodology.status set to "accepted" in statePhase 3 → Phase 4:
"status": "accepted" in statein_progress or pendingPhase 4 completion:
→ Redirect: "Let's finish evaluating the requirements first. Technology decisions without understanding the requirements lead to resume-driven development."
→ Redirect: "We're deciding the big picture right now. Component details come in Phase 3. Let's make sure the overall structure is right first."
→ Redirect: "Component [X] needs to be accepted first. Each component builds on previous decisions."
Watch for and flag these throughout ALL phases:
~30 seconds. Free. No account. Every finding cites a rule and a line of evidence.