<< All versions
Skill v1.0.0
currentTrusted Publisher100/100microsoft/cat-agent-skills/blog-content-auditor
──Details
PublishedOctober 22, 2026 at 07:31 AM
Content Hashsha256:79f6d33272dc82a0...
Git SHAb40cadab772d
──Files
Files (1 file, 3.2 KB)
SKILL.md3.2 KBactive
SKILL.md · 106 lines · 3.2 KB
version: "1.0.0" name: blog-content-auditor description: Use this skill when the user asks to audit, review, score, rationalise, prioritise, refresh, or improve blog posts, article libraries, newsletters, or thought-leadership content.
Blog Content Auditor
Use this skill to audit individual blog posts or a wider blog library and produce a prioritised improvement plan.
Scope options
The skill can handle:
- A single post supplied as text or a URL.
- A list of posts supplied by the user.
- A blog index if the agent has suitable web or content access.
- Draft posts before publication.
- Existing thought-leadership libraries that need rationalisation.
Evidence rules
- Audit only posts that are supplied or accessible.
- Do not invent traffic, ranking, engagement, conversion, recency, or SEO performance data.
- If the user wants performance analysis and no analytics are available, state that analytics were not available and assess editorial quality only.
- Quote or reference specific observed features when justifying recommendations.
Default audit dimensions
Score each post from 0 to 5 for:
- Reader promise - Is the post clear about why the reader should care?
- Opening strength - Does the opening create relevance quickly?
- Argument shape - Is there a coherent point of view, not just a list of observations?
- Evidence and examples - Are claims supported by examples, experience, references, or practical detail?
- Structure and scanability - Is the post easy to navigate?
- Distinctiveness - Does it avoid generic commentary?
- Next action - Does the post help the reader think or act differently?
Workflow
- Identify the available posts and the audit objective.
- Read each accessible post fully enough to assess the dimensions.
- Score each dimension with evidence.
- Identify content that should be kept, refreshed, merged, expanded, or retired.
- Produce a prioritised editorial backlog.
- If relevant, identify themes missing from the current library.
Single-post output
markdown
## Blog audit summaryOverall score: [x]/5Recommended action: [Keep / Refresh / Rewrite / Merge / Retire]## Scorecard| Dimension | Score /5 | Evidence | Recommended improvement ||---|---:|---|---|## Highest-value improvements1.[Improvement]2.[Improvement]3.[Improvement]## Suggested revised angle[Suggested angle]
Library output
markdown
## Blog library audit summary| Post | Overall score /5 | Recommended action | Reason ||---|---:|---|---|## Priority backlog| Priority | Post | Action | Rationale | Effort ||---|---|---|---|---|## Content gaps-[Gap]
References
This skill includes supporting reference material. Read the relevant reference file when the task needs additional structure, rubric detail, examples, or checklist support.
references/blog-audit-rubric.md- use this when additional structure, examples, or checks are useful for the task.
Quality checklist
Before responding, check:
- The audit does not claim performance data unless supplied.
- Recommendations are tied to evidence.
- The user can see what to improve first.
- The output separates editorial judgement from measurable analytics.