Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Choose the correct CLAUDE.md or LESSONS.md template for journalism projects. Use when starting a new project, setting up documentation, or unsure which template category fits best. Provides decision trees and selection guidance for 6 journalism-focused template types.
.claude/skills/jamditis-template-selector/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-03 | ✗→✓ | ▲ Improved | 72% | 0% |
| case-04 | ✗→✓ | ▲ Improved | 94% | 0% |
| case-05 | ✗→✓ | ▲ Improved | -9% | 0% |
| case-07 | ✗→✓ | ▲ Improved | 43% | 0% |
| case-08 | ✗→✓ | ▲ Improved | 35% | 0% |
Match projects to the correct template category. Don't guess, use the decision tree.
| Project type | Template category | |-------------|-------------------| | Newsroom AI tool, fact-checker | editorial-tool | | Conference, summit, workshop | event-website | | Newsletter, podcast, blog | publication | | Investigation, data journalism | research-project | | CMS workflow, syndication | content-pipeline | | Historical collection, archive | digital-archive |
Is it for journalism/publishing?
├── Yes →
│ ├── Has specific end date/event? → event-website
│ ├── Ongoing content series? → publication
│ ├── Historical/preservation? → digital-archive
│ ├── Newsroom tool/AI? → editorial-tool
│ ├── CMS/automation? → content-pipeline
│ └── Investigation with scope? → research-project
│
└── No (general software) →
Use general template or another collectionUse when: Building newsroom tools, writing assistants, fact-checkers, AI-powered research tools, moderation systems, workflow automation for journalists
Not for: Generic productivity tools, non-journalism software
Signs it's an editorial tool:
Use when: Building sites for conferences, summits, workshops, or events with specific dates
Not for: Ongoing content, organization websites without events, event management software (that's editorial-tool)
Signs it's an event website:
Use when: Building newsletters, podcasts, blogs, magazines, or ongoing content series with regular publishing schedule
Not for: One-off articles, static content sites, CMSes that power publications (that's editorial-tool)
Signs it's a publication:
Use when: Building investigative journalism projects, data journalism, analysis with defined scope and end date
Not for: Ongoing research operations, general data analysis, tools that support research (that's editorial-tool)
Signs it's a research project:
Use when: Building CMS workflows, publishing automation, content syndication systems, feed aggregators, cross-posting tools
Not for: Manual publishing workflows, single-site content management, publications themselves
Signs it's a content pipeline:
Use when: Building historical collections, preservation projects, research databases, document repositories
Not for: Active news sites, CMS platforms, search engines
Signs it's a digital archive:
| Project | Wrong choice | Right choice | Why | |---------|-------------|--------------|-----| | Newsletter dashboard | editorial-tool | publication | It's about the publication, not the tech | | AI fact-checker | research-project | editorial-tool | It's a tool, not a one-off investigation | | Event registration system | event-website | editorial-tool | It's a tool that serves events | | Podcast website | content-pipeline | publication | It's a publication, not automation | | Document search tool | digital-archive | editorial-tool | It's a tool, not the archive itself | | FOIA tracking system | research-project | editorial-tool | It's a tool that supports research |
If you're still unsure, ask:
project-memory/templates/project-retrospective/templates/claude-skills-journalism/
├── project-memory/templates/
│ ├── editorial-tool.md
│ ├── event-website.md
│ ├── publication.md
│ ├── research-project.md
│ ├── content-pipeline.md
│ └── digital-archive.md
│
└── project-retrospective/templates/
├── editorial-tool.md
├── event.md
├── publication.md
└── research-project.mdPick the specific template that fits. "General" is a last resort.
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | pass→pass | 12,895 | 3,984 | -69% | 1 | 1 | 0% | 1,596 | 2,070 | +30% | 0 | 0 | — |
case-02 | pass→pass | 8,354 | 3,218 | -61% | 1 | 1 | 0% | 1,490 | 1,919 | +29% | 0 | 0 | — |
case-03 | fail→pass | 6,933 | 3,934 | -43% | 1 | 1 | 0% | 1,188 | 2,040 | +72% | 0 | 0 | — |
case-04 | fail→pass | 6,631 | 5,212 | -21% | 1 | 1 | 0% | 1,151 | 2,237 | +94% | 0 | 0 | — |
case-05 | fail→pass | 12,938 | 3,523 | -73% | 1 | 1 | 0% | 2,272 | 2,074 | -9% | 0 | 0 | — |
case-06 | pass→pass | 7,726 | 2,205 | -71% | 1 | 1 | 0% | 1,166 | 1,733 | +49% | 0 | 0 | — |
case-07 | fail→pass | 9,565 | 4,261 | -55% | 1 | 1 | 0% | 1,513 | 2,157 | +43% | 0 | 0 | — |
case-08 | fail→pass | 9,207 | 4,069 | -56% | 1 | 1 | 0% | 1,551 | 2,100 | +35% | 0 | 0 | — |
case-09 | fail→pass | 9,235 | 4,238 | -54% | 1 | 1 | 0% | 1,557 | 2,046 | +31% | 0 | 0 | — |
case-10 | fail→pass | 10,861 | 2,844 | -74% | 1 | 1 | 0% | 1,789 | 1,797 | +0% | 0 | 0 | — |
case-11 | fail→pass | 9,690 | 2,877 | -70% | 1 | 1 | 0% | 1,687 | 1,809 | +7% | 0 | 0 | — |
case-12 | fail→pass | 9,539 | 3,053 | -68% | 1 | 1 | 0% | 1,637 | 1,797 | +10% | 0 | 0 | — |
case-13 | fail→pass | 9,786 | 3,626 | -63% | 1 | 1 | 0% | 1,505 | 1,992 | +32% | 0 | 0 | — |
case-14 | fail→pass | 9,108 | 3,338 | -63% | 1 | 1 | 0% | 1,582 | 1,924 | +22% | 0 | 0 | — |
case-15 | fail→pass | 18,584 | 2,094 | -89% | 1 | 1 | 0% | 3,414 | 1,672 | -51% | 0 | 0 | — |
case-16 | fail→pass | 9,982 | 2,520 | -75% | 1 | 1 | 0% | 1,784 | 1,801 | +1% | 0 | 0 | — |
case-17 | pass→pass | 7,189 | 3,625 | -50% | 1 | 1 | 0% | 1,168 | 1,987 | +70% | 0 | 0 | — |
case-18 | fail→pass | 7,388 | 3,133 | -58% | 1 | 1 | 0% | 1,302 | 1,831 | +41% | 0 | 0 | — |
case-19 | fail→pass | 7,880 | 2,935 | -63% | 1 | 1 | 0% | 1,386 | 1,866 | +35% | 0 | 0 | — |
case-20 | fail→pass | 38,706 | 2,689 | -93% | 1 | 1 | 0% | 1,429 | 1,809 | +27% | 0 | 0 | — |
case-21 | pass→pass | 9,852 | 3,410 | -65% | 1 | 1 | 0% | 1,534 | 1,826 | +19% | 0 | 0 | — |
case-22 | pass→pass | 11,952 | 4,383 | -63% | 1 | 1 | 0% | 1,851 | 2,079 | +12% | 0 | 0 | — |
case-23 | fail→pass | 16,348 | 3,559 | -78% | 1 | 1 | 0% | 2,653 | 1,924 | -27% | 0 | 0 | — |
case-24 | fail→pass | 8,999 | 2,363 | -74% | 1 | 1 | 0% | 1,436 | 1,662 | +16% | 0 | 0 | — |
DecimalAI ran this skill against gemini-3.6-flash twice over the same eval suite — once with the skill loaded and once without — and compared the two runs case by case. 24 cases were attempted. The headline lift of +75 percentage points is the difference between those two pass rates over the 24 comparable cases.
Without the skill loaded, the model failed this case. With it loaded, the same prompt on the same model passed. This is one improved case from the latest verified run; every case, including any that regressed, is in the table above.
| Model | Method | Date | Lift |
|---|---|---|---|
| gemini-3.6-flash | verified | 8/21/2026 | +55% |
Other measured skills in the registry, with their headline benchmark lift.