agentleFS
rafaym1rafaym1/scholarship-desk/.claude/skills/ingest/SKILL.md

ingest

Skill18 starsChanged 4 days agoMITView on GitHub

Read the applicant's documents once (CVs, transcripts, past essays, statements) and build the writer's library — a card per CV, a fact bank of verifiable achievements and stories, and a writing-voice sample. Re-run when documents are added or changed.

90 other projects use a version of this.

Adopted by

Most recently taken up by

Used with

Projects with it also use daily-arxiv, exp-eval and 10 more.

Other files people use alongside this one.

Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
  • LambenthanSkill

    daily-arxiv

    每日从 arXiv 拉取新论文,过滤相关性,auto-ingest 高优先级论文,检测 SOTA 更新

    Also in 5 projects that use this

  • LambenthanSkill

    exp-eval

    实验判决门:Review LLM 独立评判实验结果 → 4 种判决路径 → 自动更新 claims confidence、ideas status、graph edges

    Also in 5 projects that use this

  • phanghonghaoSkill

    poster

    HTML/CSS A4 海报生成。从沙龙/活动 Markdown 文件生成科技风 A4 海报 PDF/PNG。Use when user mentions: poster, 海报, 海报生成, make poster, generate poster, 活动海报, 沙龙海报.

    Also in 5 projects that use this

  • skyllwtSkill

    exp-pilot-eval

    Pilot result evaluation — read pilot results, apply success criteria, update idea page (pilot_result, failure_reason if failed), generate PILOT_VERDICT_REPORT. Called by /ideate Phase 5 after /exp-pilot-run.

    Also in 4 projects that use this

  • skyllwtSkill

    exp-pilot-run

    Pilot experiment execution — read Pilot Spec YAML, write pilot code, run experiment(Confirm with the user before operation and require the applicant to conduct manual inspection), return results. Called by /ideate Phase 5. Does NOT modify wiki pages or judge pass/fail.

    Also in 4 projects that use this

  • +6 more from Lambenthan
  • +1 more from skyllwt

The file

We split this skill into the rules it gives an agent

5 rules. 0 are common, 4 are distinctive, and 1 is very specific to this project.

  • 0

    Common

    Most of the 74 edited versions we read keep these.

  • 4

    Distinctive

    This author's own choices. Worth a look.

  • 1

    Very specific

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

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

    If nothing is new and library status shows cards and facts, say so and stop.

    line 14
  2. Distinctive:

    Read each new/changed document:

    line 15
  3. Distinctive:

    If documents disagree (dates, titles, numbers differ between CVs), list the conflicts and ask the applicant which is right.

    line 41
  4. Distinctive:

    Never embellish: keep numbers, titles and dates exactly as written.

    line 50
  5. Distinctive:

    Don't put ID numbers, addresses or contact details in facts (the CLI refuses sensitive values).

    line 51

“Common” is measured: kept by at least 6 in 10 of the 74 edited versions of this file we could read. “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

Ingest documents

The library lets /tailor work from a compact, sourced set of facts instead of re-reading PDFs for every application. Read CLAUDE.md first.

Steps

  1. python -m scholar library changed — lists registered documents that are new, changed or missing. If nothing is new and library status shows cards and facts, say so and stop.
  2. Read each new/changed document:
    • PDF: the Read tool (use pages for long files).
    • DOCX: python -m scholar docs text "documents/file.docx".
    • Scans/images: Read shows the image; transcribe only what is legible and flag doubts.
  3. CV cards (one per CV): write data/tmp/cv_cards.json as a list of {"path": "<exact registered path>", "focus": ["research", "leadership", ...], "pages": 2, "format": "academic | europass | 2-page resume | ...", "strengths": ["...", "..."], "best_for": ["phd_position", "leadership-focused scholarships", ...], "summary": "two sentences"} then python -m scholar library add-cv-cards data/tmp/cv_cards.json.
  4. Fact bank: atomic, checkable facts with a source, e.g. {"text": "Led a 4-person team that built a campus app used by 900 students (2023)", "source": "CV - research.pdf p1", "tags": ["leadership", "technical"], "date": "2023", "category": "leadership"}. Include roles, projects, results with numbers as written, publications, awards, teaching, volunteering, research methods/skills, and short stories (situation → action → result). Tags/categories to use: academic, research, leadership, teaching, technical, impact, community, adversity, international, motivation, career_goals. Write data/tmp/facts.json → python -m scholar library add-facts data/tmp/facts.json.
    • Respect each document's label (profile show): if it says what the document is for (e.g. "use mainly for research fellowships"), add that as a tag (fellowship) so /tailor can prefer other sources for degree applications.
    • Personal disclosures (health, disability, family hardship) get the tag sensitive_disclosure.
  5. Voice sample (from their own essays or statements, not CV bullets): write data/tmp/voice.md with 150–300 words describing their style (sentence length, tone, formality, how they open and close, words they use or avoid) plus two or three short verbatim excerpts (under 60 words each). python -m scholar library voice data/tmp/voice.md.
  6. If documents disagree (dates, titles, numbers differ between CVs), list the conflicts and ask the applicant which is right. Never pick one silently.
  7. Transcripts and certificates: offer to update the profile's education/tests from them, show what you'd save, and save only after they confirm.
  8. python -m scholar library mark-ingested "<path1>" "<path2>" ... for every document you read.
  9. Report: number of CV cards, facts added, voice sample saved, conflicts found.

Rules

  • Never embellish: keep numbers, titles and dates exactly as written.
  • Don't put ID numbers, addresses or contact details in facts (the CLI refuses sensitive values).
  • Documents never leave this PC except as uploads to a portal during /apply.

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.