Tessl • The management layer for AI agent skills

SkillPublic

multi-agent-patterns

Design multi-agent architectures for complex tasks. Use when single-agent context limits are exceeded, when tasks decompose naturally into subtasks, or when specializing agents improves quality.

Quality

67%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Passed

No findings from the security scan

Fix and improve this skill with Tessl

tessl review fix ./plugins/sadd/skills/multi-agent-patterns/SKILL.md

Quality

Content

38%Weight 40%Scale 1-5

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

The content is well-sectioned and covers relevant patterns with useful Claude Code implementation notes, but it is over-long, mixes in an unrelated memory-architecture treatise, and keeps everything inline with no progressive disclosure. Tightening the prose and splitting the memory material into a referenced file would materially improve it.

Suggestions

Move the 'Memory System Design' section (lines 251-371) into a separate references file and link to it from a one-line pointer; it duplicates the earlier 'Memory and State Management' section and is tangential to multi-agent patterns.

Replace the conceptual 'Why Multi-Agent Architectures' prose and the DMR benchmark table with a concise rationale; remove explanations of concepts Claude already knows (vector stores, lost-in-middle effect).

Add at least one fully executable end-to-end example (e.g., a concrete Task-tool invocation with real prompt strings) and turn the Guidelines list into a workflow with explicit validation checkpoints.

Dimension	Reasoning	Score
Conciseness	The body is noticeably verbose: it includes extended conceptual prose on context bottlenecks and a ~120-line 'Memory System Design' treatise (vector stores, knowledge graphs, a DMR benchmark table) that largely explains concepts Claude already knows and is tangential to a multi-agent patterns skill.	2 / 5
Actionability	It provides some concrete Claude Code guidance ('call specialized subagents using the Task tool', 'use shared files as the coordination mechanism') and structural templates, but many examples are high-level pseudocode outlines rather than copy-paste-ready, executable commands.	3 / 5
Workflow Clarity	A numbered Guidelines list and per-pattern 'when to use / advantages / disadvantages' give a rough sequence, but validation and feedback-loop checkpoints are only mentioned as advice ('Validate outputs before passing between agents') rather than embedded as explicit checkpoints in a workflow.	3 / 5
Progressive Disclosure	No bundle files exist and the skill is a single monolithic document with no references to separate files; the large Memory System Design section is content that clearly belongs in its own reference file but is fully inlined.	2 / 5
	Total	10 / 20 Passed

Description

78%Weight 40%Scale 1-5

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

A strong, third-person description that cleanly states both the capability and explicit use-when triggers. Its main weakness is specificity: it names one action ('design') rather than enumerating the concrete design activities the skill supports.

Suggestions

List concrete design actions (e.g., 'select a coordination pattern, partition subtasks, define handoff protocols') to lift specificity from a single action to several.

Add a common synonym or phrasing users might say (e.g., 'multi-agent systems' or 'agent orchestration') to round out trigger term coverage.

Dimension	Reasoning	Score
Specificity	The description names the domain ('multi-agent architectures') and one concrete action ('Design ... for complex tasks'), but offers only a single action rather than a list of specific capabilities, matching the 1-2 concrete actions anchor.	3 / 5
Completeness	It explicitly answers both what ('Design multi-agent architectures for complex tasks') and when ('Use when single-agent context limits are exceeded, when tasks decompose naturally into subtasks, or when specializing agents improves quality') with concrete trigger phrases.	5 / 5
Trigger Term Quality	It includes natural, relevant terms users would say ('multi-agent', 'single-agent context limits', 'subtasks', 'specializing agents') with good coverage, though a few common synonyms are missing.	4 / 5
Distinctiveness Conflict Risk	'Multi-agent architectures' is a fairly clear niche with distinct triggers and low overlap risk, though it sits adjacent to general orchestration/task-management skills, so minor overlap remains.	4 / 5
	Total	16 / 20 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 — 16 / 16 Passed

Validation for skill structure

No warnings or errors.

Repository: NeoLabHQ/context-engineering-kit
Path: plugins/sadd/skills/multi-agent-patterns/SKILL.md
Revision: 23e2428

Reviewed: 7 days ago

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.

multi-agent-patterns

Table of Contents

Is this your skill?