agenthub

Multi-agent collaboration plugin that spawns N parallel subagents competing on the same task via git worktree isolation. Agents work independently, results are evaluated by metric or LLM judge, and the best branch is merged. Use when: user wants multiple approaches tried in parallel — code optimization, content variation, research exploration, or any task that benefits from parallel competition. Requires: a git repo.

Quality

76%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

—

The risk profile of this skill

Fix and improve this skill with Tessl

tessl review fix ./engineering/agenthub/skills/agenthub/SKILL.md

Quality

Content

65%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 body is well-organized and actionable with concrete commands and a clear lifecycle, but two of three reference files are unlinked with their content inlined, and the destructive merge workflow lacks an explicit validation checkpoint. Tightening both would lift the content score.

Suggestions

Add a validation checkpoint before /hub:merge — e.g. 'Verify the winning branch passes tests/build in its worktree before merging' — to satisfy the destructive-operation feedback-loop requirement.

Move the inlined DAG Model and coordination-pattern content into references/dag-patterns.md and references/coordination-strategies.md and link them one level deep from the body instead of duplicating.

Drop or trim the "When This Skill Activates" trigger-phrase list since the same triggers already appear in the description's "Use when" clause, to reduce token redundancy.

Dimension	Reasoning	Score
Conciseness	The body is mostly lean — command tables, templates, and the coordinator protocol earn their tokens — but the "When This Skill Activates" trigger list duplicates the description's "Use when" clause and a few table descriptions could be trimmed; efficient with minor over-explanation.	4 / 5
Actionability	Provides concrete executable commands with flags (e.g. `python scripts/result_ranker.py --session {id} --eval-cmd "pytest bench.py --json" --metric p50_ms --direction lower`), slash-command and script tables, and a post-format template; minor gaps remain (external `clawhub install`, placeholder `{id}` tokens).	4 / 5
Workflow Clarity	The INIT→DISPATCH→MONITOR→EVALUATE→MERGE lifecycle is clearly sequenced with a state-transition table, but the destructive/batch merge step (merge winner, `git tag hub/archive/{session}/agent-{i}` losers, worktree cleanup) has no explicit validation checkpoint before merge, so workflow clarity is capped at 3 per the destructive-operations rule.	3 / 5
Progressive Disclosure	Templates are cleanly signaled ("Templates are defined in `references/agent-templates.md`"), but `coordination-strategies.md` and `dag-patterns.md` exist in the bundle yet are never linked from the body — their content (DAG model, coordination patterns) is inlined instead, fitting the 'some structure, content that should be separate is inline' anchor.	3 / 5
	Total	14 / 20 Passed

Description

88%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 description: it states concrete capabilities across the full lifecycle and gives explicit trigger guidance in third person. The only weakness is a slightly broad catch-all clause and a few missing natural trigger synonyms.

Dimension	Reasoning	Score
Specificity	Lists multiple concrete actions — "spawns N parallel subagents competing on the same task via git worktree isolation", "results are evaluated by metric or LLM judge", "the best branch is merged" — covering the full lifecycle comprehensively.	5 / 5
Completeness	Explicitly answers both what (spawn/evaluate/merge competing agents) and when ("Use when: user wants multiple approaches tried in parallel — code optimization, content variation, research exploration...") with concrete trigger phrases.	5 / 5
Trigger Term Quality	Includes natural phrases like "multiple approaches tried in parallel", "parallel competition", "code optimization", and "content variation", but misses a few common synonyms a user might say; fits the 'good coverage, a few terms missing' anchor better than the comprehensive anchor.	4 / 5
Distinctiveness Conflict Risk	The worktree-based parallel-competition mechanism is a clear niche with minimal conflict risk, but the closing "any task that benefits from parallel competition" is broad enough to create minor overlap with single-agent optimization skills.	4 / 5
	Total	18 / 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: alirezarezvani/claude-skills
Path: engineering/agenthub/skills/agenthub/SKILL.md
Commit: 19392f7

Reviewed: ~3 hours 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.