agentic-os

Build persistent multi-agent operating systems on Claude Code. Covers kernel architecture, specialist agents, slash commands, file-based memory, scheduled automation, and state management without external databases. Use when building a persistent multi-agent system on Claude Code with its own memory, commands, and scheduling.

Quality

83%

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

Quality

Content

78%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.

Highly actionable architecture guide with excellent copy-paste templates and no concept padding. Its main weaknesses are length-driven: a monolithic single-file layout that should offload detail to references, and some redundancy in the rationale and command examples.

Suggestions

Move the platform-specific scheduling examples (LaunchAgent plist, systemd timer, pm2 config) into a references/ file such as references/scheduling.md, keeping only one representative example inline — the three full configs are the largest token block in the body.

Split the full kernel CLAUDE.md template and agent definition template into reference files (e.g., references/kernel-template.md, references/agent-template.md), leaving a skeleton plus the key principle in SKILL.md.

Remove the '// data/projects/website-v2.json' comment from the JSON code fence — it is invalid JSON if copied verbatim — and cut either the Key Principle or Over-Engineered Routing statement of the 'inspectable and editable' rationale, which appears twice.

Dimension	Reasoning	Score
Conciseness	No tutorial padding — it never explains what systemd, git, or JSON are — and the examples are load-bearing templates. Minor trimmable fat keeps it at 4 rather than 5: the 'inspectable and editable' rationale is stated twice (Key Principle and Over-Engineered Routing), and the Standard Command Set table lists commands (/outreach, /apply-jobs, /analytics) that are never defined.	4 / 5
Actionability	Fully concrete, copy-paste-ready templates throughout: a complete kernel CLAUDE.md, agent definition format, LaunchAgent plist, systemd unit + timer pair, pm2 ecosystem config, and JSON state schemas. One trivial blemish ('// data/projects/...' comment inside the JSON fence, which breaks if pasted verbatim) does not drop it below the 5 anchor's 'specific examples cover the common cases'.	5 / 5
Workflow Clarity	Routing is a clean numbered 1–5 sequence, commands have step lists, and the reflection pattern gives an explicit feedback loop. It stays at 4 rather than 5 because the overall 'assemble this OS' setup sequence is implied by the layer tour instead of being an explicit ordered checklist, and there is no validation checkpoint for verifying the built system works.	4 / 5
Progressive Disclosure	The skill is monolithic: no bundle files exist, and content that clearly belongs in separate reference files — three full platform scheduling configs (plist, systemd, pm2), the entire kernel template, and the agent template — is inlined in a 380+ line body. Section headers are good (above the 2 anchor), but 'content that should be separate is inline' matches 3; the under-50-line exception does not apply.	3 / 5
	Total	16 / 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: concrete, comprehensive capability list with an explicit and specific 'Use when...' clause. Trigger coverage is good but omits several natural synonyms that appear only in the body's activation list.

Dimension	Reasoning	Score
Specificity	The description enumerates the skill's full capability surface concretely: 'kernel architecture, specialist agents, slash commands, file-based memory, scheduled automation, and state management without external databases'. This matches the 5 anchor (multiple specific concrete capabilities, comprehensive coverage) rather than 4, which is reserved for lists with minor gaps.	5 / 5
Completeness	It explicitly answers both questions: what ('Build persistent multi-agent operating systems on Claude Code. Covers kernel architecture, specialist agents, slash commands, file-based memory, scheduled automation, and state management') and when ('Use when building a persistent multi-agent system on Claude Code with its own memory, commands, and scheduling'). Matches the 5 anchor with concrete trigger conditions; not 4 since the 'when' clause is already explicit and specific.	5 / 5
Trigger Term Quality	'persistent multi-agent system', 'memory, commands, and scheduling', and 'Claude Code' give good natural keyword coverage, but common synonyms a user might actually say — 'agentic OS', 'personal OS', 'agent coordinator' — are absent. This fits 4 ('good keyword coverage; a few natural terms missing'), not 5's comprehensive synonym coverage.	4 / 5
Distinctiveness Conflict Risk	The niche is clear (persistent, memory-owning multi-agent systems on Claude Code), but 'multi-agent' and 'scheduling' overlap with generic agent-orchestration and automation skills. Minor overlap risk with closely related skills fits 4; the distinct trigger set keeps it above 3.	4 / 5
	Total	18 / 20 Passed

Validation

93%

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 — 15 / 16 Passed

Validation for skill structure

Criteria	Description	Result
metadata_version	'metadata.version' is missing	Warning

	Total	15 / 16 Passed

Repository: affaan-m/ECC
Path: skills/agentic-os/SKILL.md
Commit: c70874f

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.