agentleFS
Sign inStart free
zhangzejingzhangzejing/ThinkReader/skills/report-skill/SKILL.md

abzu-report

Skill42 starsChanged 2 months agoMITView on GitHub

Create a concise evidence-linked ThinkReader paper report with original figures and explicit boundaries. Use for /report only.

The file

We split this skill into the rules it gives an agent

4 rules. About 0 are common, 2 are distinctive, and 2 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈2

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 2

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Read prompts/flash.md for a fast report, prompts/pro.md for deep mechanism/evidence analysis, or both when mode is automatic.

    line 13
  2. Very specific:

    Read rules.md, tools.md, writer.md and data/quality.json.

    line 14
  3. Distinctive:

    Read the App-provided report context, PaperMap and relevant full sections before writing.

    line 15
  4. Distinctive:

    Follow important methods, equations, figures and experiments through their complete section context.

    line 29

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

ThinkReader Report Skill

/report produces the lightest research artifact: a short, continuous explanation that lets a peer understand the paper's idea, evidence and limits in a few minutes. It is derived output, never Reader truth.

Required reading order

  1. Read prompts/base.md.
  2. Read prompts/flash.md for a fast report, prompts/pro.md for deep mechanism/evidence analysis, or both when mode is automatic.
  3. Read rules.md, tools.md, writer.md and data/quality.json.
  4. Read the App-provided report context, PaperMap and relevant full sections before writing.

File ownership

  • prompts/base.md owns synthesis and prose quality.
  • mode prompts own only depth differences.
  • rules.md owns report shape, evidence, figures, language and formulas.
  • tools.md owns assumed context, evidence, figure, validation and output capabilities.
  • data/quality.json owns deterministic quality floors.
  • writer.md owns proposal, validation and commit flow.

Non-negotiable method

  • Explain the causal story, not the table of contents and not a rewritten abstract.
  • Follow important methods, equations, figures and experiments through their complete section context.
  • Cite only labels and numeric claims that resolve to current evidence.
  • Let the LLM decide which evidence matters and write the report; use tools for exact extraction, figure staging, deep links, file assembly and validation.
  • Make uncertainty and missing evidence visible rather than filling gaps.

Boundaries

Write only through controlled output tools under output/report/<paper_id>/. Existing annotations are optional, not a prerequisite. After commit the App first appends the report tag to an existing annotation covering the cited evidence; preserve its content, author, marks, hints, replies and layout. Keep uncovered evidence in the report citations; do not create PDF cards for evidence bookkeeping. Repeated runs must not multiply cards or tags; matching and tag writes are tool work, not extra LLM output. Never modify RAW/ or Intelligence. Do not restore legacy context JSON, cleaner output, HTML Reader or public CLI flows.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

Check your setup, free