Skip to content
SkillScript-led7,509 tokens in use

SR Screener

Protocol-driven screening for systematic, scoping and rapid reviews: two blinded AI reviewers, a third-reviewer adjudicator, QC checks and PRISMA 2020 counts.

57,881 stars on the repository Custom / UnknownImbad0202/academic-research-skills · sr-screener
/plugin marketplace add Imbad0202/academic-research-skills /plugin install academic-research-skills
View source

Install command as written in the repository's README

Skill label

measured from its files

What installing SR Screener costs

259tokens idle
7,509tokens in use
25,912tokens, up to

Idle is what sits in context while the skill is installed. In use is its instructions, loaded when it runs. Up to adds the files its instructions link. Counts are approximate, made offline with cl100k_base.

Model dependence Script-led

Script-ledMixedModel-led

Most of this skill is scripts, so it behaves much the same whichever model runs it.

This describes the skill's structure, how much of it is scripts. It does not say how well it works. Scripts: scripts/build_outputs.py, scripts/build_workflow.py, scripts/merge_decisions.py, scripts/prepare_fulltext.py, scripts/prepare_records.py, scripts/srlib.py, templates/workflows/ft_screening.template.js, templates/workflows/ta_screening.template.js.

Touches

shelldetectednetworkdetectedwrites filesdetected

Declared tags come from the skill's frontmatter; detected ones from its scripts, its code blocks and the commands it names in inline code. This is detected, not audited.

How it scores

Quality

How well the skill is written, counted from its own files: its frontmatter, its steps, examples, stated output and constraints, its size and whether every file it ships is one it names.How we score it →
Good
  • Frontmatter8/15
  • Structure25/25
  • Examples15/15
  • Output10/15
  • Size6/15
  • Resources15/15

Security

Starts at 100 and loses points for each thing in the skill's files that a reader should look at first.How we score it →
Good
  • Runs shell commands−5

    does this on the machine, and the frontmatter does not say so

    SKILL.md, line 256
  • Writes files−8

    does this on the machine, and the frontmatter does not say so

    SKILL.md, line 256
  • Reaches the network−10

    does this on the machine, and the frontmatter does not say so

    scripts/build_outputs.py, line 375

Repository health

Pre-computed score based on development activity, maintenance, community, maturity, and trend momentum.How we score it →
Excellent
  • Development activity100
  • Maintenance100
  • Community60
  • Maturity32
  • Momentum40

SR Screener turns a proposal or PROSPERO protocol into confirmed eligibility rules, then screens titles, abstracts and full texts with two independent AI reviewers. Disagreements go to a third reviewer, and every exclusion carries an ordered code and a short reason.

It reads RIS, PubMed .nbib, Web of Science and CSV exports, removes duplicates and runs in resumable batches. The outputs are an Excel log, RIS groups for EndNote or Zotero, PRISMA counts, a methods draft and a corpus file that hands off to the academic-paper skill. People keep the final sign-off.

What You Get

  • A confirmed screening protocol and config file built from your proposal
  • Screening decisions with one ordered exclusion code and a short reason per record
  • An Excel log, RIS groups for EndNote or Zotero, and PRISMA 2020 counts with a flow diagram
  • A methods paragraph with an AI-use statement and a literature corpus file for the writing stage

Common Use Cases

  • Title and abstract screening of a database export for a systematic review
  • Piloting eligibility criteria against a team’s own labelled records
  • Adjudicating screening conflicts from a Rayyan or Covidence export
  • Auditing exclusions and reporting the selection process for PRISMA

What Makes It Unique

The skill enforces a fixed order of work. The protocol must be confirmed before any record is read, a pilot must pass against the team’s labels before the full run, a cost estimate needs a clear yes, and a sample of records both reviewers excluded gets a senior recheck. It reports observed agreement, kappa and PABAK together, and states that a single model reading twice is not two independent reviewers.

Under The Hood

The skill loads a long SKILL.md with eight modes and four agent files. Deterministic steps run as local Python scripts that parse exports, merge decisions and build outputs; they use the standard library, with openpyxl optional for the Excel log. Reviewers run as lean subagents with read and search tools only, so they cannot browse, write files or see each other’s work. The skill reads your exports and PDFs and writes working files, logs and reports to your project folder. Text inside records is treated as data and never as instructions.

Join founders building with open source

Opinionated takes, migration guides, cost-saving tips, and insights from the open source ecosystem.

Subscribe on SubstackJoin 750+ subscribers