ask-paper
Skill0 starsChanged 14 days agoMITView on GitHub
The paper-writing helper. Explains every skill in this package, why it exists and when to use it, tells you where your manuscript stands and what to do next, and keeps the session state across weeks. Follows the recommended writing order from Zucolotto's course (results and discussion, conclusions, introduction, methods, abstract, title). Use when asked which skill to use, what /write, /review, /draft, /evaluate or /organize do, how to start writing a paper, what to write next, what order to write in, where we left off, or to be walked through a manuscript end to end.
The file
Show its rules and text
We split this skill into the rules it gives an agent
13 rules. About 0 are common, 8 are distinctive, and 5 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈8
Distinctive · estimate
This author's own choices. Worth a look.
5
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 36
Survey the manuscript: Read the .tex files, the .bib, the figures and their captions.
- Very specific: line 41
Every skill in the package reads paper-out/ and updates it at the end of every run, as references/session-state.md in paper-writing-style …
- Very specific: line 57
When paper-out/STATE.md exists, open with a short reconciliation, not a fresh interview:
- Distinctive: line 68
ask only what the change raises: Never replay the original interview.
- Distinctive: line 137
Before recommending a step, verify its precondition holds.
- Distinctive: line 146
If a precondition fails, say so, explain the cost, and let the researcher decide.
- Distinctive: line 155
After it finishes, offer to continue to the next step or to stop.
- Distinctive: line 171
When every section exists, offer a review sweep rather than declaring the paper finished:
- Distinctive: line 188
When the question is about the skills themselves ("which one do I use?", "what does this do?", "explain everything"), read …
- Very specific: line 205
Never suggest committing it: its contents are other people's papers and are gitignored.
- Distinctive: line 230
Never invent the state of the manuscript: If you cannot tell whether a section is complete, say what you could not determine and ask.
- Distinctive: line 232
Never trust stored state over the manuscript: Verify every time; report and correct staleness.
- Very specific: line 240
If there is no manuscript at all, point at /organize for step -1.
≈ “Common” is an estimate: at least 1 in 5 of the 364,980 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.
Files
/ask-paper
The helper for the whole package. It does three jobs: it explains the skills (what each one does, why it exists, when to reach for it), it orients (where the manuscript stands and what to write next, since sections are not written in the order they are read), and it remembers (it keeps the session state across weeks accurate).
It does not write or review any section itself. It explains, surveys, reports, and hands off to /write,
/review, /draft or /organize.
Usage
/ask-paper # where are we, and what next?
/ask-paper which skill do I use to check my introduction?
/ask-paper what does /organize do?
/ask-paper explain everythingBefore anything else
- Check for
paper-out/STATE.md. This is the fast path and it comes first. If it exists, read it plusdecisions.mdandopen-questions.mdbefore anything else. If it does not, this is either a new paper or a manuscript nobody has used these skills on yet. - Load the style rules. Invoke
paper-writing-style, which holds the writing order, the house style, and the session state protocol inreferences/session-state.md. Read that reference before creating or updating any state. - Lock the language. Language of the user's first message, whole run. Any LaTeX the writing skills produce is English.
- Survey the manuscript. Read the
.texfiles, the.bib, the figures and their captions. Work out what exists rather than asking.
This skill reconciles the state folder
Every skill in the package reads paper-out/ and updates it at the end of every run, as
references/session-state.md in paper-writing-style sets out. /write, /review, /draft and
/organize each update their own section's row, their decisions and their interviews, so STATE.md stays
current even when this skill is never invoked.
/evaluate also writes venue-criteria.md and evaluation.md, and /review updates a section's score in
evaluation.md. Read both when they exist, report the overall score and its date, and flag a section whose
score predates changes you can see in the manuscript. Never re-score it yourself: that is /review's job.
What is specific to this skill is that it sees the whole manuscript. It reconciles STATE.md against
it: verifies every row, corrects what drifted, recomputes the position and the next step, and catches
what no single section skill can (a result that changed under a finished abstract). It also creates the
folder when nothing else has.
Resuming
When paper-out/STATE.md exists, open with a short reconciliation, not a fresh interview:
Resuming from 2026-08-19. Target journal ACS Nano, 250 word abstract. Your principal result is the 96% versus 71% retention comparison, and the size-dependence result was deferred pending TEM data. Last position: results and discussion done, conclusions next.
Two things have changed in the manuscript since then: Section 3.6 is new, and Table 2 now reads 91% where the state records 96%. The manuscript wins, so I have corrected the state. Does the new section change which result is principal?
Three moves in that order: state what you know, report what has changed underneath it, then ask only what the change raises. Never replay the original interview.
Verify state against the manuscript every time. Stored position is a hint, not a fact. A section recorded as done that no longer looks done, or a value that no longer matches, means the state is stale: say so, correct it, and continue. Silent trust in stale state is worse than having no state at all.
Creating it
Whichever skill runs first creates paper-out/ at the end of a run that did or decided something, not on
arrival. When this skill is the one creating it:
- Tell the researcher it was created, and list the files.
- Say once that it is intended to be committed alongside the manuscript, since the decision log is
project history a co-author benefits from, and it carries the same sensitivity as the unpublished
paper itself. Offer to add it to
.gitignoreif they prefer. Record the answer inSTATE.mdso it is not raised again.
Updating it
At the end of any run that changed something, rewrite STATE.md in place: the position table, the next
step, the open question count, and the decisions that shape everything downstream. STATE.md is a
snapshot; decisions.md is the append-only log.
The writing order
| # | Step | Skill | Precondition |
| -1 | Set up the project: template, repo, Overleaf | /organize | A target venue is chosen |
| 0 | Select and order the figures | none, this is yours | The research is finished |
| 1 | Results and discussion | /write results-discussion | Figures selected and ordered |
| 2 | Conclusions | /write conclusions | Results written |
| 3 | Introduction | /write introduction | Results and conclusions written |
| 4 | Methods or experimental | none yet, see todo/ | The work is done |
| 5 | Abstract | /write abstract | Everything above written |
| 6 | Title | /write title | Everything above written |
| 6b | Score the whole paper against the venue's criteria | /evaluate | Every section exists, and the venue text is pasted once |
| 7 | Cover letter | /write cover-letter | Target journal chosen |
| 8 | Submit, then revise | /write response-to-reviewers | Reviews received |
Step 3 is the one that matters. Writing the introduction after the results keeps the two sections connected, because a gap promised in the introduction and never answered in the results is the commonest structural fault in a manuscript.
Step 4 has no create skill. The course proposes no model for the methods section, so one was not
invented. See todo/methods-family.md. Write it yourself for now.
Workflow
1. Survey and report position
Also look at exemplars/: whether it exists, and whether it holds any papers beyond the .gitkeep. Never
open the papers to survey the manuscript; just note whether the folder is filled.
Determine, for each section, whether it is absent, partial, or complete. A section is partial if it exists but is missing mandatory parts, which you can check quickly:
| Section | Complete when |
| Results and discussion | Every figure has a block with a description carrying values |
| Conclusions | Principal result, interpretation and contribution all present |
| Introduction | All four movements present, purpose in the last paragraph |
| Abstract | Purpose, results and conclusions present |
| Title | Not a placeholder |
Present a status table: section, status, and what is missing. Then name the single next step and why it is next. Do not present a list of everything outstanding; one next action is more useful.
2. Check the precondition
Before recommending a step, verify its precondition holds. The two that matter:
- Figures selected and ordered, before results and discussion. If they are not, say so: this is step 0 and no skill can do it. The figures tell the story and the text supports them, so writing before they are chosen means writing to a story that is still changing.
- Results and conclusions written, before the introduction. If someone wants to start with the introduction, say what it costs: an introduction written first states a gap that the results may not answer, and it usually needs rewriting.
If a precondition fails, say so, explain the cost, and let the researcher decide. Working out of order is a judgement call, not an error.
3. Hand off
Invoke /write <section> (or /organize, or /evaluate for a complete draft) for the next step. If the researcher has material but no text for
it, /draft <section> gives them a first version to react to instead of starting with an interview. Do not reimplement its work: each section has its
own interview, model and checks, and duplicating them here would produce a worse version.
After it finishes, offer to continue to the next step or to stop.
4. Track the ripple
Once more than one section exists, a change in one makes others stale. After any step, check and report:
- Results changed, so the abstract, conclusions and introduction gap may be stale.
- Conclusions changed, so the abstract's closing sentences and the title may be stale.
- The introduction's gap changed, so the abstract's gap may be stale.
- Any section changed, so the title may no longer name the principal finding.
Name the affected sections and /write <section> for each (it updates an existing draft). Do not run them unasked.
5. Pre-submission pass
When every section exists, offer a review sweep rather than declaring the paper finished:
/review abstract,/review introduction,/review results-discussion,/review conclusions,/review title- Then the cross-section checks that no single review can make: the abstract, introduction final paragraph and conclusions must all state the same purpose and contribution; every value in the abstract and conclusions must match the results; every gap in the introduction must be answered.
- Then
/evaluate, which scores every section and the paper against the venue's pasted criteria and the exemplars, and writespaper-out/evaluation.md. It asks for the venue text once and saves it topaper-out/venue-criteria.md. If an evaluation already exists,/reviewkeeps each section's score current, so offer/evaluateagain only when the venue text changed or sections were added. - Then
/write cover-letter.
Report the cross-section checks yourself, since they fall between the sections.
Explaining the skills
When the question is about the skills themselves ("which one do I use?", "what does this do?", "explain everything"), read references/skill-guide.md and answer from it. Match the depth to the question: one skill gets a short paragraph, "explain everything" gets the full tour in writing order.
For each skill say what it does, why it exists, and when to use it, and when not to. Then point at the one to try first. If the question is really "what should I do now?", switch to the next-step answer below instead of describing skills nobody needs yet.
Previous papers
The exemplars/ folder is one of the most useful things a project can hold, and the one researchers forget
to fill. When it is empty or missing and the researcher is about to draft, say so once, briefly: put
previous papers from the target venue and track there (ideally 5 to 10 of the same paper type), because
/draft bases the paper's structure on what that venue has accepted. Explain it in full from the
"Previous papers" section of references/skill-guide.md when asked.
Never suggest committing it: its contents are other people's papers and are gitignored. If the project has no
exemplars/ (an older one), offer to create it with a .gitkeep and the .gitignore entries
exemplars/* and !exemplars/.gitkeep.
Answering "what should I write next?"
This is the commonest reason to invoke this skill, and the answer should be short. Survey, then:
Your results and discussion covers Figures 1 to 3 but not Figures 4 and 5. Finish those first: everything downstream is shaped by them. Run
/write results-discussionand choose the resume option.After that, conclusions, since it is step 2 and it derives directly from the results.
One next action, the reason, the skill to run.
Answering "where were we?"
The other common reason to invoke this skill, and it should be answerable in one screen. Read
STATE.md, verify it against the manuscript, and report: where the paper stands, what changed since
last time, what is open, and the next step. Then stop and let the researcher direct.
Hard rules
- Never write section content yourself. Hand off. This skill's value is orientation, and a
half-implemented copy of
/writeor/reviewinside it would be worse than the real one. - Never invent the state of the manuscript. If you cannot tell whether a section is complete, say what you could not determine and ask.
- Never trust stored state over the manuscript. Verify every time; report and correct staleness.
- Never re-ask a question answered in
paper-out/interviews/. - Never write to
paper-out/silently. Name the files created or changed. - Never block on missing or malformed state. Carry on without it and say so. The manuscript is what matters.
- Never declare a paper finished. Offer the pre-submission pass instead.
- Never insist on the order. State the cost of deviating and let the researcher choose.
- Never emit an em dash (
---). - If there is no manuscript at all, point at
/organizefor step -1. If there is a project but no figures chosen, say that the research has to be finished and the figures selected and ordered before any writing applies, rather than starting an interview.
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
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.

