architecture-decision

Document an architectural decision — record the decision and rationale, retain into Hindsight, and update CLAUDE.md if needed.

Quality

58%

Does it follow best practices?

Impact

—

No eval scenarios have been run

Securityby

Passed

No known issues

Optimize this skill with Tessl

npx tessl skill review --optimize ./skills/architecture-decision/SKILL.md

Quality

Content

77%

Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.

This is a well-structured skill with a clear multi-step workflow, concrete templates, and appropriate validation checkpoints (user approval gates). Its main weaknesses are minor verbosity—the diagram complexity callout and some explanatory notes could be trimmed—and the inline ADR template could potentially be externalized. Overall it provides actionable, executable guidance for documenting architectural decisions.

Suggestions

Trim the diagram complexity callout block or move it to a shared reference; it adds cognitive overhead for what is an optional sub-step.

Remove the 'Important' bullet about constraints vs decisions—Claude already understands this distinction and it wastes tokens.

Dimension	Reasoning	Score
Conciseness	Generally efficient but includes some unnecessary elaboration. The diagram complexity callout at the top feels tangential to the core ADR workflow. Some sections like the 'Important' notes explain things Claude would already understand (e.g., 'if there were no alternatives, it's not really a decision — it's a constraint').	2 / 3
Actionability	Provides a concrete ADR template with specific fields, clear step-by-step instructions for each phase (parse, structure, retain, check CLAUDE.md, confirm), and a specific confirmation output format. The guidance is specific enough to be directly executable.	3 / 3
Workflow Clarity	The 6-step workflow is clearly sequenced with explicit checkpoints: user review/approval at step 3, conditional CLAUDE.md update requiring user approval at step 5, and a confirmation summary at step 6. The feedback loop of presenting the ADR for review before retention is a good validation checkpoint.	3 / 3
Progressive Disclosure	The content is reasonably well-structured with clear sections, but the full ADR template is inline rather than referenced. For a standalone skill with no bundle files, this is acceptable, but the diagram complexity rating block at the top feels like it belongs in a shared reference rather than being embedded here. References to render-diagram and Hindsight skills are mentioned but not linked.	2 / 3
	Total	10 / 12 Passed

Description

40%

Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.

The description identifies a clear niche around architectural decision records with specific tooling references (Hindsight, CLAUDE.md), giving it good distinctiveness. However, it lacks an explicit 'Use when...' clause and misses common trigger terms like 'ADR' or 'design decision' that users would naturally use. The description would benefit from expanded trigger terms and explicit selection guidance.

Suggestions

Add a 'Use when...' clause such as 'Use when the user asks to document, record, or capture an architectural decision, ADR, design decision, or technical choice.'

Include common trigger term variations like 'ADR', 'architecture decision record', 'design decision', 'technical decision', 'why we chose'

Add more specific actions such as 'creates a structured ADR with context, options considered, decision outcome, and consequences'

Dimension	Reasoning	Score
Specificity	Names the domain (architectural decisions) and some actions (record decision/rationale, retain into Hindsight, update CLAUDE.md), but the actions are somewhat high-level and could be more concrete about what 'documenting' entails specifically.	2 / 3
Completeness	Describes what it does (document an architectural decision, record rationale, retain into Hindsight, update CLAUDE.md) but lacks any explicit 'Use when...' clause or trigger guidance. Per rubric guidelines, a missing 'Use when' clause caps completeness at 2, and the 'when' is entirely absent here, warranting a score of 1-2. The description reads more like a task summary than selection guidance.	1 / 3
Trigger Term Quality	Includes relevant terms like 'architectural decision', 'rationale', 'Hindsight', and 'CLAUDE.md', but misses common variations users might say such as 'ADR', 'architecture decision record', 'design decision', 'why did we choose', or 'technical decision'.	2 / 3
Distinctiveness Conflict Risk	The combination of architectural decisions, Hindsight retention, and CLAUDE.md updates creates a very specific niche that is unlikely to conflict with other skills. The mention of specific tools (Hindsight, CLAUDE.md) makes it clearly distinguishable.	3 / 3
	Total	8 / 12 Passed

Validation

100%

Warnings & errors only

Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.

Validation — 11 / 11 Passed

Validation for skill structure

No warnings or errors.

Repository: AndreJorgeLopes/devflow
Commit: 8d872bf

Reviewed: 1 day ago

Table of Contents

Discovery Implementation Validation

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.