haowjy/prose-analysis
.agents/skills/prose-analysis/SKILL.md
Mechanical prose metrics — sentence length, opener variety, dialogue ratio, repetition, pronoun distribution. Use when you need quantitative signals about prose before making subjective judgments, or when comparing a draft against the project's baseline.
data-ai
Updated Apr 16, 2026
$ install --global
skillsauthnpx skillsauth add haowjy/pokemon-amber prose-analysisInstall this skill globally with one command. Works with Claude Code, Cursor, and Windsurf.
Security Scan Results
3 of 9 scanners reported clean
Some scanners were skipped, did not run, or reported a non-clean status. Review each row below.
3
Scanners Passed
9
Scanners in report
Clean
TrivyContainer and dependency vulnerability scanner
95%
Clean
SemgrepStatic code analysis for vulnerabilities
95%
Clean
mcp-scan (Snyk)Model Context Protocol security validation
95%
Skipped
Snyk (dep)Open source security scanning
50%
Skipped
Socket.devSupply chain security analysis
50%
Skipped
VirusTotalMulti-engine malware detection
50%
Skipped
CrowdStrikeAdvanced threat intelligence
50%
Skipped
OSV-ScannerOpen Source Vulnerability database check
50%
Skipped
OWASP Dep-Check
50%
Last scanned: Apr 24, 2026, 8:54 PM1.8s1 file scanned
SKILL.md
- name:
- prose-analysis
- description:
- Mechanical prose metrics — sentence length, opener variety, dialogue ratio, repetition, pronoun distribution. Use when you need quantitative signals about prose before making subjective judgments, or when comparing a draft against the project's baseline.
Prose Analysis — Quantitative Signals
Scripts that measure mechanical properties of prose. These produce numbers, not quality verdicts. The numbers become useful only when compared against the project's own baseline — research shows no reliable universal thresholds for "good" prose exist.
What the Scripts Measure
The bundled resources/analyze.sh script produces:
- Sentence length distribution — mean, median, min, max, standard deviation. Compare across chapters for consistency. High variance often indicates varied rhythm (usually good); low variance may indicate monotone pacing.
- Sentence opener variety — categorizes first words of sentences (pronoun, article, conjunction, name, other). Heavy pronoun-leading ("I did... I saw... I felt...") is a common POV drift signal.
- Dialogue-to-narration ratio — proportion of lines that are dialogue vs narration. Useful for scene-type comparison (action scenes should skew narration, conversation scenes should skew dialogue).
- Repetition detection — words and short phrases that appear multiple times within a sliding window of N paragraphs. Catches echoed words the author didn't intend.
- Pronoun distribution — counts of I/me/my vs he/she/they/etc. Useful as a POV consistency check (a first-person chapter with sudden third-person pronouns needs investigation).
Running the Script
bash resources/analyze.sh <file.md> [window_size]
file.md— the markdown file to analyzewindow_size— optional, number of paragraphs for the repetition detection window (default: 5)
Output is plain text with labeled sections. Standard unix tools only (awk, sed, grep, wc, sort).
What the Scripts Don't Measure
Some metrics from the research literature are excluded because they're either unreliable, require Python NLP libraries, or don't produce actionable signals for creative writing:
- Readability scores (Flesch-Kincaid, etc.) — not a quality predictor for fiction
- Vocabulary richness (raw TTR) — methodologically weak at varying text lengths. MATTR is better but requires windowed computation that's impractical in pure bash.
- Perplexity — requires a language model to compute, and very low perplexity just means "predictable," not "bad"
- Adverb/filter word density — modest effect sizes, heavily genre-confounded
Resources
resources/analyze.sh— the analysis scriptresources/antipatterns.md— AI writing antipatterns, honestly categorized as research-backed vs community folklore. Read this when reviewing AI-generated drafts and you want to know which signals are worth investigating vs which are unreliable.resources/baseline.md— how to establish a project baseline from existing chapters and compare new drafts against it. Read this before interpreting analysis output for the first time on a project.
Related Skills
haowjy/writing-staffing
data-ai
VerifiedTrustedCommunity
Team composition for writing workflows — which agents to spawn, how many, what focus areas to assign, and how to scale effort. Use when composing critic panels, dispatching researchers, staffing draft/revise loops, or setting up brainstorm fan-outs.
SKILL.mdUpdated Apr 16, 2026
haowjy/writing-principles
testing
VerifiedTrustedCommunity
What fiction readers actually want, framed as four composable reward channels (transportation, aesthetic, social simulation, flow), and the specific documented ways alignment training damages each one. Grounded in reader-psychology research and empirical NLP findings. Load when drafting prose, critiquing a draft, deciding whether to show or tell, diagnosing why a passage feels flat, or reasoning about why a scene is or isn't working.
SKILL.mdUpdated Apr 16, 2026
haowjy/writing-issues
testing
VerifiedTrustedCommunity
Logging and referencing writing issues — craft problems, tics, inconsistencies, and structural concerns found during analysis, critique, or review. Use when an agent identifies something worth tracking beyond a single critique report: repeated tics across chapters, inconsistencies that affect multiple scenes, structural problems that need the author's attention, or patterns that should be fixed in revision.
SKILL.mdUpdated Apr 16, 2026
haowjy/writing-artifacts
development
VerifiedTrustedCommunity
Shared artifact convention between orchestrators — what goes where in `$MERIDIAN_FS_DIR/` and `$MERIDIAN_WORK_DIR/`, how artifacts flow between phases, and what each directory means. Use whenever work artifacts, style files, knowledge entries, drafts, or critique reports are being created, referenced, or discussed.
SKILL.mdUpdated Apr 16, 2026