task-observer

Monitors task execution for skill improvement opportunities. Use this skill during ANY multi-step task, agentic workflow, or substantive work session where the agent is using tools and producing deliverables. It captures patterns, user corrections, workflow insights, and methodology worth preserving as reusable skills. Also triggers during post-task feedback discussions and when the user explicitly mentions skill observations, improvements, the observation log, skill taxonomy, or asks the agent to watch for skill opportunities. Also known as "One Skill to Rule Them All" — trigger on this phrase too. IMPORTANT: this skill should be invoked at the start of every task-oriented session — if you are about to use tools to produce deliverables, invoke this skill first. For reliable activation, pair this description with a CLAUDE.md instruction or harness-level session-start hook (see Recommended Activation Setup) — description-level matching alone is not enforceable.

Quality

—

Does it follow best practices?

Impact

—

No eval scenarios have been run

Securityby

Passed

No known issues

Quality

Content

62%

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

The skill is highly actionable and its workflows are clearly sequenced with strong validation feedback loops, but it is markedly verbose and ships as one ~1500-line monolith with no bundle files, undercutting its own 'Lean Content' guidance. Actionability and workflow clarity are excellent; conciseness and progressive disclosure pull the score down.

Suggestions

Cut the non-load-bearing rationale ('Why This Skill Exists', the Pre-Flight narrative, the multi-license catalogue, and the self-narrating intro paragraphs) to honour the skill's own Lean Content rule — keep rules, anti-patterns, formats, and worked shell, drop the backstory.

Split the bulk into one-level-deep bundle files (e.g. references/review-procedure.md for the Step 0–8 review, references/templates.md for the observation/handoff/log templates, references/confidentiality.md for the five-layer sweep) and leave a concise overview plus signposts in SKILL.md.

Deduplicate the confidentiality framing, which is restated across the Taxonomy, Confidentiality Safeguards, and Quick Reference sections — consolidate to one authoritative pass.

Dimension	Reasoning	Score
Conciseness	The body runs ~1500 lines with heavy rationale and self-narrating prose ('Why This Skill Exists', 'The Pre-Flight Principle' narrative), repeated confidentiality framing, a multi-license explainer, and a lengthy 'Lean Content' section that itself violates the lean-content rule — matching the verbose anchor rather than the 'mostly efficient' level above.	1 / 3
Actionability	It supplies concrete, executable shell (the grep/sed observation-numbering, pre-write assertion, and post-write TOCTOU renumber snippets), exact file paths, and copy-paste templates (observation format, handoff doc, log structure), meeting the fully-executable anchor.	3 / 3
Workflow Clarity	Multi-step processes are explicitly sequenced with validation checkpoints: the Session Start Protocol, the Step 0–8 Comprehensive Review, and the check-then-act-then-verify numbering flow with pre-write assertion and post-write renumber feedback loops, plus a self-enforcement checklist.	3 / 3
Progressive Disclosure	It is a single monolithic file with no bundle files in references/, scripts/, or assets/; section headers and external GitHub doc links provide structure, but content that should be split out (templates, the full review procedure, the license catalogue) is inlined, matching the 'some structure but content that should be separate is inline' anchor.	2 / 3
	Total	9 / 12 Passed

Description

82%

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 explicit about both capability and triggers and packs in many natural keywords, but it is written in second person (incurring the specificity penalty) and its trigger is so broad that it competes with almost every other skill. It is strong on completeness and trigger coverage but weak on distinctiveness.

Suggestions

Rewrite the description in third person ('Monitors...', 'Invokes...') to avoid the second-person specificity penalty — replace 'if you are about to use tools... invoke this skill first' with a third-person formulation.

Narrow the activation language toward skill-improvement intent (e.g. 'Use when the session may surface reusable skill patterns or methodology worth capturing') so it is less likely to fire on every generic task alongside other skills.

Trim the meta-commentary about CLAUDE.md hooks and enforceability out of the description; that setup guidance belongs in the body, keeping the description focused on what/when.

Dimension	Reasoning	Score
Specificity	It names multiple concrete actions ('Monitors task execution for skill improvement opportunities', 'captures patterns, user corrections, workflow insights, and methodology worth preserving as reusable skills'), which would warrant a 3, but it uses second-person voice ('if you are about to use tools to produce deliverables, invoke this skill first'), triggering the rubric's -1 specificity penalty.	2 / 3
Completeness	It explicitly answers both what ('Monitors task execution... captures patterns, user corrections, workflow insights') and when ('Use this skill during ANY multi-step task, agentic workflow, or substantive work session... Also triggers during post-task feedback discussions'), with explicit 'Use when'-style triggers.	3 / 3
Trigger Term Quality	It covers a broad set of natural phrases a user would say ('multi-step task', 'agentic workflow', 'post-task feedback', 'skill observations', 'improvements', 'observation log', 'skill taxonomy', 'One Skill to Rule Them All'), matching the rubric's good-coverage anchor rather than the partial-coverage anchor below.	3 / 3
Distinctiveness Conflict Risk	Its trigger condition is 'ANY multi-step task, agentic workflow, or substantive work session where the agent is using tools and producing deliverables' — broad enough to overlap with nearly every task-oriented skill, and the description itself concedes 'description-level matching alone is not enforceable'; it is not the level-3 'clear niche' anchor but is more purposeful than the level-1 generic example.	2 / 3
	Total	10 / 12 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
skill_md_line_count	SKILL.md is long (1525 lines); consider splitting into references/ and linking	Warning

	Total	15 / 16 Passed

Repository: rebelytics/one-skill-to-rule-them-all
Commit: 379a8a3

Reviewed: 4 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.