agent-release-manager

Agent skill for release-manager - invoke with $agent-release-manager

2.51x

Quality

—

Does it follow best practices?

Impact

78%

2.51x

Average score across 3 eval scenarios

Securityby

Advisory

Suggest reviewing before use

Quality

Content

27%

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

The body is a sprawling, redundant release-management playbook: it contains genuinely concrete tool calls but buries them in duplicated examples and placeholder/pseudocode, with no validation feedback loops for risky batch operations and no progressive disclosure structure.

Suggestions

Collapse the duplicated v1.0.72 examples into a single canonical walkthrough and remove the buzzword prose to tighten the token budget.

Replace placeholder strings and fix the corrupted path separators ('$' → '/') so commands are actually executable, and add explicit 'validate → on failure fix → retry' checkpoints around version bumps and deploys.

Move stable reference material (semantic-versioning strategy, CI/CD integration, metrics) into reference files under references/ and link to them one level deep from the overview.

Dimension	Reasoning	Score
Conciseness	The ~370-line body is padded with redundant content — the v1.0.72 changelog and PR body each appear twice — plus buzzword prose ('seamless version management', 'comprehensive testing') that adds tokens without new guidance.	1 / 3
Actionability	It provides concrete MCP calls, gh CLI commands, and a GitHub Actions YAML block, but many examples are placeholders ('[updated package.json]', '[comprehensive release description]') and path separators are corrupted ('$' instead of '/'), making them non-copy-paste-ready.	2 / 3
Workflow Clarity	The release pipeline is sequenced (prepare, test, PR, review, merge) with a TodoWrite checklist, but destructive batch operations (multi-package version bumps, deploys) lack explicit validation checkpoints and 'if it fails, then…' feedback loops.	2 / 3
Progressive Disclosure	It is a monolithic single file with no references/, scripts/, or assets/ bundle and no external navigation; reference-style material (strategies, CI/CD config, metrics) that should be split out is all inline.	1 / 3
	Total	6 / 12 Passed

Description

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 frontmatter description is minimal and meta: it identifies the skill as an agent skill for release-manager but states no concrete capabilities and provides no usage triggers. It fails to tell Claude what the skill does or when to invoke it.

Suggestions

Replace the meta phrasing with concrete actions, e.g. 'Coordinates multi-package releases: bumps versions, runs validation suites, and opens release pull requests.'

Add an explicit 'Use when...' trigger clause naming natural user terms such as 'release', 'deploy', 'publish', 'version bump', or 'changelog'.

Drop the internal invocation syntax ('$agent-release-manager') from the description; it is not a term a user would naturally say.

Dimension	Reasoning	Score
Specificity	The description 'Agent skill for release-manager - invoke with $agent-release-manager' names only the domain and offers no concrete actions; it is a meta-statement about the skill rather than what it does.	1 / 3
Completeness	It answers neither 'what' (only vaguely — 'agent skill for release-manager') nor 'when' (no 'Use when...' clause or equivalent trigger guidance).	1 / 3
Trigger Term Quality	Beyond the single token 'release-manager', there are no natural trigger terms a user would say (e.g. release, deploy, version, publish); 'invoke with $agent-release-manager' is internal jargon.	1 / 3
Distinctiveness Conflict Risk	'release-manager' denotes a specific niche unlikely to overlap with unrelated skills, but the description is too generic to establish clearly distinct triggers, so it is not a 3.	2 / 3
	Total	5 / 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 — 16 / 16 Passed

Validation for skill structure

No warnings or errors.

Repository: ruvnet/claude-flow
Commit: 03d4f84

Reviewed: 2 days 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.