autonomous-agents

Autonomous agents are AI systems that can independently decompose goals, plan actions, execute tools, and self-correct without constant human guidance. The challenge isn't making them capable - it's making them reliable. Every extra decision multiplies failure probability.

Quality

—

Does it follow best practices?

Impact

—

No eval scenarios have been run

Securityby

Advisory

Suggest reviewing before use

Quality

Content

35%

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

The body is content-rich but token-heavy: it re-explains known concepts, wraps code in non-fenced quote blocks that hurt executability, and bundles everything into one monolithic file without progressive disclosure. Tightening, fencing the code, and splitting deep material into referenced files would substantially improve it.

Suggestions

Move the duplicated description and the explanatory prose about what autonomous agents / LLMs are into a short intro only; assume Claude's competence and keep only the non-obvious reliability guidance inline.

Wrap every code example in fenced code blocks (```python) instead of bare triple-quote strings, and make examples self-contained or clearly mark the helper functions they assume, so guidance is executable rather than illustrative.

Split the bulk into one-level-deep reference files (e.g. PATTERNS.md for ReAct/Plan-Execute/Reflection implementations, GUARDRAILS.md for safety patterns, VALIDATION.md for the checks) and keep SKILL.md as a concise overview that links to them, adding explicit validate→fix→retry checkpoints in the destructive/batch workflows.

Dimension	Reasoning	Score
Conciseness	The body restates the frontmatter description verbatim, then explains what an autonomous agent is and spends lines on concepts Claude already knows ('LLMs are trained to be helpful', quadratic cost intuition), padding a 1080-line document with unnecessary context — matching the verbose/knows-concepts anchor.	1 / 3
Actionability	It provides substantial code, but the examples are wrapped in bare triple-quote blocks rather than fenced code blocks so they won't render or run as code, and many rely on undefined helpers (generator.generate, summarize, estimate_cost, planner.plan_next), making them illustrative rather than copy-paste executable.	2 / 3
Workflow Clarity	The 'Sharp Edges' entries are consistently sequenced (Situation → Symptoms → Why → Fix) giving a clear structure, but the multi-step agent workflows lack explicit validation/feedback checkpoints (validate → fix → retry) and the recommended fixes read as loose bullet lists rather than gated steps, capping clarity at 2 for these batch/destructive operations.	2 / 3
Progressive Disclosure	No bundle files exist, so the skill is a single 1080-line monolith organized only by headers; the sectioning gives some structure, but content that should be split out (framework catalog, full pattern implementations, validation checklist) is inlined with no one-level-deep references, matching the 'some structure, content that should be separate is inline' anchor.	2 / 3
	Total	7 / 12 Passed

Description

67%

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 is specific and distinct, but it is missing the explicit 'Use when…' trigger guidance that anchors a top completeness score; that guidance is relegated to the body. Adding a trigger clause with natural user keywords would lift completeness and trigger-term quality.

Suggestions

Append an explicit 'Use when...' clause to the description naming natural user phrases (e.g. autonomous agent, agent loop, ReAct pattern, LangGraph, agentic AI, goal decomposition) so the 'when' is stated in the description itself, not just the body.

Tighten the opener so the description leads with capability actions rather than the 'challenge/reliability' framing, which is context rather than a trigger.

Confirm the description uses third person throughout (it does) and keep it concise — avoid pulling the full 'When to Use' keyword list verbatim, a representative subset suffices.

Dimension	Reasoning	Score
Specificity	Lists multiple concrete actions: 'independently decompose goals, plan actions, execute tools, and self-correct without constant human guidance', matching the score-3 anchor of several specific concrete actions.	3 / 3
Completeness	It clearly answers 'what' (the agent capabilities), but the 'when to use it' guidance is not in the description — it lives only in the body's '## When to Use' section — so the 'when' is implied rather than explicit per the anchor that caps completeness at 2 when the trigger clause is missing.	2 / 3
Trigger Term Quality	Includes some relevant terms ('autonomous agents', 'self-correct', 'decompose goals') but the description itself omits the natural user-facing keywords (langgraph, react pattern, agentic ai, agent loop) that only appear later in the body's 'When to Use' list, so coverage is incomplete.	2 / 3
Distinctiveness Conflict Risk	It carves a clear niche (reliability of autonomous agents) with distinct framing ('making them reliable', 'every extra decision multiplies failure probability') that is unlikely to trigger for adjacent skills like memory-systems or tool-building.	3 / 3
	Total	10 / 12 Passed

Validation

87%

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 — 14 / 16 Passed

Validation for skill structure

Criteria	Description	Result
skill_md_line_count	SKILL.md is long (1085 lines); consider splitting into references/ and linking	Warning
frontmatter_unknown_keys	Unknown frontmatter key(s) found; consider removing or moving to metadata	Warning

	Total	14 / 16 Passed

Repository: boisenoise/skills-collections
Commit: f36337d

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.