Skip to content

kgraph57/paper-writer-skill

paper-writer

Verified

Medical/scientific paper writing workflow skill. Manages the full pipeline from literature search to submission-ready manuscript. Creates and manages a project directory with IMRAD-format section files, literature matrix, reference management, and quality checklists. Supports both English and Japanese papers. Triggers: 'write paper', 'paper-write', 'start manuscript', '論文を書く', '論文執筆', '論文プロジェクト', 'manuscript', 'research paper', '原稿作成'.

FileKindSize
SKILL.mdskill md77.7 KB
examples/case-studies/discovery-to-draft.mdmarkdown5.3 KB
references/abstract-formats.mdmarkdown5.1 KB
references/adversarial-review.mdmarkdown21.6 KB
references/ai-disclosure.mdmarkdown6.4 KB
references/ai-for-science-model.mdmarkdown12.8 KB
references/citation-guide.mdmarkdown2.4 KB
references/citation-verification.mdmarkdown10.5 KB
references/clinical-trial-registration.mdmarkdown4.7 KB
references/coauthor-review.mdmarkdown14.2 KB
references/coi-detailed.mdmarkdown6.8 KB
references/desk-rejection-prevention.mdmarkdown6.1 KB
references/hook-compatibility.mdmarkdown6.1 KB
references/humanizer-academic.mdmarkdown12.0 KB
references/imrad-guide.mdmarkdown2.5 KB
references/journal-reformatting.mdmarkdown5.6 KB
references/journal-selection.mdmarkdown13.5 KB
references/keywords-guide.mdmarkdown4.6 KB
references/master-reference-list.mdmarkdown11.3 KB
references/multilingual-guide.mdmarkdown12.4 KB
references/novelty-check.mdmarkdown13.8 KB
references/open-access-guide.mdmarkdown4.6 KB
references/pubmed-query-builder.mdmarkdown16.7 KB
references/reporting-guidelines-full.mdmarkdown10.6 KB
references/reporting-guidelines.mdmarkdown3.2 KB
references/section-checklist.mdmarkdown4.5 KB
references/statistical-reporting-full.mdmarkdown8.6 KB
references/statistical-reporting.mdmarkdown22.5 KB
references/submission-portals.mdmarkdown4.1 KB
references/supplementary-materials.mdmarkdown3.9 KB
references/tables-figures-guide.mdmarkdown7.7 KB
references/word-count-limits.mdmarkdown4.6 KB
scripts/analysis-template.pyexecutablescript19.6 KB
scripts/compile-manuscript.shexecutablescript6.9 KB
scripts/forest-plot.pyexecutablescript16.9 KB
scripts/sr-dedup.pyexecutablescript12.5 KB
scripts/sr-pdf-link.pyexecutablescript6.8 KB
scripts/sr-prisma-count.pyexecutablescript8.7 KB
scripts/table1.pyexecutablescript9.7 KB
scripts/word-count.shexecutablescript12.8 KB
templates/abstract.mdmarkdown1.3 KB
templates/acknowledgments.mdmarkdown1.7 KB
templates/analysis-workflow.mdmarkdown5.1 KB
templates/case-abstract.mdmarkdown1.2 KB
templates/case-introduction.mdmarkdown1.7 KB
templates/case-report.mdmarkdown2.2 KB
templates/conclusion.mdmarkdown1.0 KB
templates/cover-letter.mdmarkdown1.3 KB
templates/data-management.mdmarkdown5.7 KB
templates/declarations.mdmarkdown6.1 KB
templates/discussion.mdmarkdown1.8 KB
templates/graphical-abstract.mdmarkdown2.7 KB
templates/highlights.mdmarkdown3.8 KB
templates/human-loop-ledger.mdmarkdown9.4 KB
templates/introduction.mdmarkdown1.8 KB
templates/limitations-guide.mdmarkdown6.2 KB
templates/literature-matrix.mdmarkdown1.2 KB
templates/methods.mdmarkdown1.9 KB
templates/preregistration.mdmarkdown19.5 KB
templates/project-init-case.mdmarkdown3.2 KB
templates/project-init.mdmarkdown3.6 KB
templates/proof-correction.mdmarkdown2.5 KB
templates/research-question.mdmarkdown18.6 KB
templates/response-to-reviewers.mdmarkdown12.5 KB
templates/results.mdmarkdown1.5 KB
templates/review-article.mdmarkdown11.6 KB
templates/revision-cover-letter.mdmarkdown10.3 KB
templates/sr-data-extraction.mdmarkdown9.2 KB
templates/sr-grade.mdmarkdown14.7 KB
templates/sr-outline.mdmarkdown13.5 KB
templates/sr-prisma-flow.mdmarkdown12.9 KB
templates/sr-prospero.mdmarkdown15.9 KB
templates/sr-rob.mdmarkdown18.3 KB
templates/sr-screening-pipeline.mdmarkdown12.1 KB
templates/study-design.mdmarkdown28.5 KB
templates/submission-ready.mdmarkdown2.5 KB
templates/title-page.mdmarkdown3.5 KB

templates/preregistration.md

View on GitHub
# Pre-Registration Template
## Phase −1 Integrity Gate — Freeze Before the 📊 Data Gate

---

## 1. Why Pre-Register

An AI-accelerated research loop is a HARKing machine unless you lock it. The
garden-of-forking-paths problem — where hundreds of plausible analysis choices
produce whatever p-value you need — already existed in human-speed science.
Add a co-scientist running model selection, covariate permutations, and outcome
re-operationalizations at machine speed (§5.1 of `references/ai-for-science-model.md`)
and the multiverse expands by orders of magnitude. The only discipline that closes
the gate is to **freeze the hypotheses and the exact primary analysis plan before
any outcome data are examined** — in a timestamped, immutable, public record.
After that lock, pre-registered analyses are confirmatory; everything else is
exploratory and must be labeled as such. Without this gate, a fast loop produces
confident slop, indistinguishable from Sakana AI Scientist v2's output.

---

## 2. Placement in the Research Loop

```
Phase −1: DISCOVERY
  ├── research question          (templates/research-question.md)
  ├── novelty check              (references/novelty-check.md)
  ├── study design + power       (templates/study-design.md)  ← input to this file
  └── 🔒 PRE-REGISTRATION LOCK   ← YOU ARE HERE
          │  (plan is now frozen; timestamp obtained)
          │
  📊 DATA GATE  ← data collection begins here (or, for retrospective data,
                   outcome data examination begins here — see §9)
          │
  Phases 0–3: analysis (bound to the frozen plan) → draft (IMRAD)
```

**Stop rule:** Data are not analyzed for outcomes until the registration record is
submitted and the timestamp is obtained. The timestamp is the evidence of the lock.

---

## 3. Which Registry

| Study type | Registry | Mandatory vs. recommended | Timing |
|---|---|---|---|
| Interventional clinical trial (Japan) | **jRCT** (jrct.niph.go.jp) | **Mandatory** under 臨床研究法 (Clinical Trials Act, 2018) for specified / non-specified clinical studies using unapproved drugs/devices or off-label use. See `references/clinical-trial-registration.md`. | Before first enrollment |
| Interventional clinical trial (Japan, UMIN legacy) | **UMIN-CTR** (umin.ac.jp/ctr/) | Recommended; widely accepted by ICMJE journals. Being superseded by jRCT for new trials. | Before first enrollment |
| Interventional clinical trial (international) | **ClinicalTrials.gov** | Mandatory for ICMJE submission; required by US law for applicable trials | Before first enrollment |
| Systematic review / meta-analysis | **PROSPERO** (crd.york.ac.uk/prospero/) | Strongly recommended; see `templates/sr-prospero.md` for the full template — do not duplicate here | Before screening begins |
| Observational / cross-sectional / cohort / case-control | **OSF Registries** (osf.io/registries) | Recommended | Before data collection; if retrospective, before outcome examination |
| Prediction model / secondary data analysis | **OSF** or **AsPredicted** (aspredicted.org) | Recommended | Before outcome examination |
| Purely exploratory / hypothesis-generating | No registry required | — | N/A |
| Animal / preclinical | **preclinicaltrials.eu** | Recommended | Before experiment |

> Japan-specific note: 臨床研究法 requires jRCT registration for 特定臨床研究
> (specified clinical research). Observational studies and retrospective chart
> reviews are typically outside its scope, but confirm with your IRB.

---

## 4. The Pre-Registration Record Template

Copy this block into your OSF / AsPredicted / UMIN-CTR / jRCT form.
Complete every field before submitting. Fields marked **(required)** have no
acceptable blank.

---

```markdown
## Pre-Registration Record

### A. Administrative

- **Title** (identical to the manuscript working title): {{PREREG_TITLE}}
- **Registry**: {{REGISTRY_NAME}}
- **Registration number** (assigned after submission): {{REG_NUMBER}}
- **Registration timestamp** (UTC): {{TIMESTAMP}}
- **Corresponding PI**: {{PI_NAME}} <{{PI_EMAIL}}>
- **Co-investigators**: {{COINVESTIGATORS}}
- **Study design input document**: templates/study-design.md, version {{VERSION}},
  dated {{DESIGN_DATE}}

---

### B. Hypotheses  **(required)**

State directionally. Vague hypotheses cannot be confirmed or disconfirmed.

- **H0 (null)**: {{H0 — e.g., "There is no difference in 30-day readmission
  rate between Group A and Group B."}}
- **H1 (alternative, directional if possible)**: {{H1 — e.g., "Group A has a
  lower 30-day readmission rate than Group B."}}
- **Direction of expected effect**: [ ] one-sided  [ ] two-sided  Reason: {{REASON}}

For secondary hypotheses, list separately as H2, H3, etc., each with own H0/H1.

**Estimand variant — descriptive / estimation primaries.** Not every primary aim is a test. For a descriptive/estimation primary (a proportion or mean you are *measuring*, not comparing against a null), do NOT force a directional H0/H1 — the estimate IS the result. Instead state:

- **Estimand**: {{the quantity being estimated — e.g., "proportion of caregivers who consulted a generative-AI tool before the visit, in the full valid-response set"}}
- **Precision target**: {{the CI half-width you are powering for — e.g., "95% Wilson CI half-width ≤ ±4.5 pp at an expected p ≈ 0.30, N = 400" (see study-design.md §5.0)}}
- **H0**: *Not applicable — estimation, not testing.*

A study may register **two descriptive co-primaries this way** without a multiplicity penalty (see §D carve-out); each is justified by its own precision target, not by a shared α.

---

### C. Study Design  **(required)**

- **Design**: {{RCT / prospective cohort / retrospective cohort / cross-sectional
  / case-control / other}}
- **Setting**: {{SETTING}}
- **Planned sample size**: N = {{N}}  (see power calculation in §F)
- **Data-collection start date** (must be AFTER lock): {{START_DATE}}
- **Data-collection end date**: {{END_DATE}}
- **Inclusion criteria**: {{INCLUSION}}
- **Exclusion criteria**: {{EXCLUSION}}

---

### D. Primary Outcome  **(required — ONE only, unless the descriptive carve-out applies)**

> Rule: if you have more than one primary outcome, you are running more than one
> study. Pick one. Pre-register the others as secondary.
>
> **Carve-out — descriptive co-primaries (no multiplicity penalty).** Purely
> descriptive co-primaries (proportions/means reported with a CI) are allowed
> **without** a multiplicity penalty **when neither is tested against a null** —
> they are estimation targets, not tests, so there is no shared type-I error to
> split (mirrors `templates/study-design.md` §3 and §5.0; use the §B estimand
> variant for each). This holds only while every co-primary stays descriptive: if
> any is compared against a reference value or against the other, it becomes a test
> and the ONE-only rule plus §H multiplicity correction re-apply.
> **Require an explicit one-line justification here**, e.g.: "Two descriptive
> co-primaries (AI-consultation prevalence; pre-visit information-source mix); each
> is an estimand with its own precision target (§B); neither is tested against a
> null, so no multiplicity penalty is taken."

- **Outcome name**: {{OUTCOME_NAME}}
- **Operationalization** (exactly what is measured, how, by whom, at what time
  point; precise enough that a stranger could reproduce it):
  {{OPERATIONALIZATION — e.g., "30-day all-cause readmission to any hospital,
  ascertained by electronic medical record review at day 30 ± 2 days post-discharge,
  confirmed by a blinded research assistant using ICD-10 codes. Counted as binary
  event (yes/no)."}}
- **Source / instrument**: {{SOURCE}}
- **Timing**: {{TIMING}}

---

### E. Secondary Outcomes

List each with the same precision as the primary. These will be reported with
multiplicity correction (§H).

| # | Outcome | Operationalization | Timing |
|---|---|---|---|
| S1 | {{NAME}} | {{OPERATIONALIZATION}} | {{TIMING}} |
| S2 | {{NAME}} | {{OPERATIONALIZATION}} | {{TIMING}} |

---

### F. Sample Size and Stopping Rule  **(required)**

- **Effect size assumed**: {{EFFECT_SIZE}} (source: {{CITE OR RATIONALE}})
- **Power**: {{POWER — e.g., 80%}}
- **Alpha**: {{ALPHA — e.g., 0.05, two-sided}}
- **Calculated N**: {{N_PER_ARM}} per arm / {{N_TOTAL}} total
- **Attrition assumed**: {{ATTRITION%}}; inflated N: {{N_INFLATED}}
- **Stopping rule** (for RCTs or sequential designs): {{STOPPING_RULE or "Not
  applicable — fixed sample design."}}
- **Software / function used**: {{e.g., R pwr::pwr.t.test(); G*Power 3.1}}

---

### G. Primary Analysis Plan  **(required — fully specified)**

> This is the single most important field. A stranger with access to the dataset
> must be able to run this analysis from this description alone.

- **Statistical model**: {{e.g., "Logistic regression"}}
- **Outcome variable**: {{VARIABLE_NAME_IN_DATASET}}
- **Primary predictor / treatment variable**: {{VARIABLE_NAME}}
- **Adjustment set / covariates** (list each, with justification):
  1. {{COVARIATE_1}} — {{REASON}}
  2. {{COVARIATE_2}} — {{REASON}}
- **Test / estimand**: {{e.g., "Wald test on the treatment coefficient; odds ratio
  with 95% CI"}}
- **Significance threshold**: α = {{ALPHA}}, {{one-sided / two-sided}}
- **Software and package version**: {{e.g., R 4.4.0, glm(); Stata 18, logistic}}
- **Handling of the analysis**: {{e.g., "intention-to-treat primary; per-protocol
  sensitivity"}}

---

### H. Multiplicity Correction  **(required if >1 outcome)**

- **Method**: {{e.g., Bonferroni, Benjamini-Hochberg FDR, hierarchical testing}}
- **Applied to**: {{primary only / all pre-specified outcomes / subgroup tests}}
- **Adjusted threshold for secondary outcomes**: α' = {{ALPHA_ADJUSTED}}

**Regime — "Exploratory analytic surface, no correction."** For a pilot whose
adjusted/association analyses are all EPV-limited and exploratory (study-design.md
§5.5), the honest move is **no multiplicity correction — because nothing here is
confirmatory**. To use this regime without it becoming a fishing licence:

- **Pre-specify a CLOSED list** of which exploratory tests may even appear in the
  abstract or conclusion. Anything not on this list cannot surface there.

| # | Exploratory test (closed list) | May appear in abstract/conclusion? |
|---|---|---|
| X1 | {{e.g., AI-use × caregiver age, adjusted Firth OR}} | Yes — as hypothesis-generating only |
| X2 | {{e.g., delayed-care count by source}} | No — supplementary table only |

- **None is a "finding."** State in the record: *"No exploratory result is reported
  as a finding; all are hypothesis-generating and require prospective replication."*
- **Estimate + CI only; suppress p.** Prefer reporting the estimate with its CI and
  **omit the p-value** for exploratory tests, so no reader (or co-scientist) can
  back-door a significance claim out of an uncorrected surface.

This regime pairs with the §D descriptive carve-out: the descriptive co-primaries
are confirmatory-by-precision, and *everything else* lives on this uncorrected
exploratory surface. The split is enforced by the Firewall table (§5).

---

### I. Missing-Data Plan  **(required)**

- **Expected missingness**: {{RATE%}} in {{VARIABLE}}
- **Mechanism assumption**: {{MCAR / MAR / MNAR; rationale}}
- **Primary missing-data approach**: {{e.g., "Multiple imputation by chained
  equations (MICE), m = 20 imputations, R mice package"}}
- **Sensitivity analysis**: {{e.g., "Complete-case analysis as sensitivity; pattern
  mixture model for MNAR sensitivity"}}

---

### J. Pre-Specified Subgroup Analyses

> Only subgroups with a plausible biological/clinical rationale should be
> pre-specified. Fishing is what the EXPLORATORY section is for.

| Subgroup | Variable | Cut-point | Rationale | Test for interaction? |
|---|---|---|---|---|
| {{SUBGROUP_1}} | {{VARIABLE}} | {{CUT}} | {{RATIONALE}} | Yes / No |
| {{SUBGROUP_2}} | {{VARIABLE}} | {{CUT}} | {{RATIONALE}} | Yes / No |

Subgroup analyses that do not appear in this table are **exploratory**.

---

### K. Exploratory Analyses (declared in advance)

List analyses that will be conducted but are explicitly labeled hypothesis-generating.
These cannot yield confirmatory claims regardless of their p-values.

- {{EXPLORATORY_1 — e.g., "Association between X and Y in the full cohort,
  unadjusted; for hypothesis generation only."}}
- {{EXPLORATORY_2}}

**Confirmatory / Exploratory Firewall (drop-in — fill one column per analysis class).**
This anti-HARKing table binds each analysis to what it is *allowed to claim* before
any data are seen. Columns = the descriptive co-primary (§D carve-out) vs. the
uncorrected exploratory surface (§H regime). Fill it as part of the frozen record:

| Row | Descriptive co-primary | Exploratory analytic surface |
|---|---|---|
| **What it is** | Estimand (proportion/mean + CI); estimation, not testing | Adjusted/association model on an EPV-limited set |
| **Inference allowed** | Point estimate with precision; "we estimate p = X (95% CI …)" | "Pattern consistent with …; hypothesis-generating only" |
| **p-value use** | None — no null is tested | **Suppressed** — estimate + CI only |
| **CI use** | Primary inferential output (Wilson) | Reported, but not as evidence of an effect |
| **Primary finding in abstract?** | Yes | No — only the closed-list items (§H), labeled hypothesis-generating |
| **Causal claim?** | No (descriptive) | No — association only |
| **Requires replication?** | No (precision pre-met) | Yes — always, prospectively |

> This is the per-analysis-class twin of the project-level table in §5; keep the two
> consistent. Any analysis that cannot be placed in one of these columns before the
> lock is, by definition, exploratory.

---

### L. Data-Collection Confirmation  **(required)**

- [ ] Data collection has NOT yet begun (prospective)
- [ ] Data exist but outcome variables have NOT yet been examined (retrospective —
  see §9 caveat)
- Date of this declaration: {{DATE}}
- Declared by: {{NAME, ROLE}}
```

---

## 5. The Confirmatory / Exploratory Firewall

| Dimension | Pre-registered (confirmatory) | Not pre-registered (exploratory) |
|---|---|---|
| **What it is** | Analysis specified in the frozen record before data examination | Anything added after the lock, regardless of reason |
| **Inference allowed** | "We reject H0 (p = X). This supports [directional claim]." | "This is a hypothesis-generating observation. Replication required." |
| **CI / p-value use** | Standard frequentist or Bayesian inference | Report descriptively; no claims of significance |
| **Manuscript placement** | Results §, primary table | Results §, secondary table, clearly labeled "Exploratory" |
| **Can it be a primary finding?** | Yes | No |
| **Requires replication?** | No (confirmatory if plan was well-powered) | Yes, always |
| **Journal disclosure** | Cite registration number | Disclose as post-hoc / exploratory |

**Hard rule:** Any analysis not in the frozen plan is exploratory, full stop. This
includes: (a) post-hoc covariate additions; (b) outcome redefinition; (c) subgroups
not in §J; (d) sensitivity analyses not pre-specified; (e) models suggested by
patterns in the data. None of these are forbidden — they are scientifically
valuable. They just cannot carry a confirmatory claim.

---

## 6. Deviation Handling

Deviations from the registered plan are expected; hiding them is the problem.

**Rule:** Every deviation is disclosed in the Methods section. The registered plan
and the actual analysis are compared side-by-side. Undisclosed deviations are
misconduct.

### Deviations Log Template

Keep this file alongside your analysis code (e.g., `deviations-log.md` in the
project repo):

```markdown
## Deviations from Pre-Registered Plan

| # | Date | Section deviated from | Planned | Actual | Reason | Impact on inference |
|---|---|---|---|---|---|---|
| D1 | {{DATE}} | §G Primary analysis | {{PLANNED}} | {{ACTUAL}} | {{REASON}} | {{e.g., "Conservative; increases Type II error"}} |
| D2 | {{DATE}} | §D Primary outcome | {{PLANNED}} | {{ACTUAL}} | {{REASON}} | {{e.g., "Outcome now composite; increases power, changes estimand"}} |
```

**Methods section disclosure template:**

```
The pre-registered analysis plan ({{REGISTRY}}, {{REG_NUMBER}}) was followed with
the following deviations: [1] {{DEVIATION_1 — what changed, why, direction of
potential bias}}. [2] {{DEVIATION_2}}. The original pre-registered analyses are
reported in Supplementary Table S{{N}} for comparison.
```

---

## 7. Registered Reports

A Registered Report (RR) is the strongest form of pre-registration: the journal
reviews and accepts the Introduction + Methods in principle **before data
collection**, guaranteeing publication regardless of the result. This structurally
eliminates publication bias. Pursue an RR when:

- The study is powered for a null result that would be scientifically important
- The research question is contentious and you anticipate publication resistance
- The team is willing to pre-commit publicly to a design that may not yet be fully
  optimized

List of journals offering RRs: cos.io/rr. For clinical journals in Japan, BMJ Open
and PLOS ONE are accessible options with RR tracks.

---

## 8. Verification Checklist (Before the 📊 Data Gate)

Complete this before any outcome data are examined or analyzed.

- [ ] Registration submitted to the correct registry (§3 decision table)
- [ ] Timestamp obtained and recorded in the project README
- [ ] Registration number inserted into the manuscript shell (Methods)
- [ ] Primary outcome: exactly ONE, fully operationalized (§D)
- [ ] Primary analysis: a stranger could run it from §G alone
- [ ] Adjustment set listed with rationale (§G)
- [ ] Sample size calculation documented with assumptions and software (§F)
- [ ] Multiplicity correction specified if >1 outcome (§H)
- [ ] Missing-data plan documented (§I)
- [ ] Pre-specified subgroups listed with interaction tests declared (§J)
- [ ] Exploratory analyses declared as such in §K
- [ ] Data collection has not started, OR retrospective-data caveat applies (§9)
- [ ] Deviations log file created and committed to project repo (initially empty)
- [ ] Human-loop ledger updated: pre-registration step logged
  (`templates/human-loop-ledger.md`)

---

## 9. Retrospective-Data Caveat

Clinicians frequently use already-collected chart data. Be honest about what
pre-registration can and cannot do here.

| Condition | What to do | What to disclose |
|---|---|---|
| Data exist; outcome variables **not yet examined** | Register before opening outcome columns. This is the strongest feasible version for retrospective data. | "Data were collected prior to registration. Outcome variables were not examined until after registration ({{REG_NUMBER}}, {{TIMESTAMP}})." |
| Data exist; some descriptive / non-outcome exploration done | Register before any outcome analysis. | "Exploratory descriptive analysis was performed prior to registration; no outcome associations were examined." |
| Data exist; outcome data already partially examined | Registration is still better than none, but is weaker. All pre-examination work is exploratory. | "This is a post-hoc analysis. The analysis plan was registered after initial data inspection ({{REG_NUMBER}}). All findings should be considered hypothesis-generating and require prospective replication." |
| Prospective data; collection not started | Gold standard. Register before enrollment. | Cite registration number; state data collection start date post-timestamp. |

**Do not claim a retrospective registration is equivalent to prospective
pre-registration.** Reviewers and readers will check the timestamp against the
data-collection end date. Honesty here is non-negotiable for a clinician publishing
under their own name.
writing-skillsobra/superpowersUse when creating new skills, editing existing skills, or verifying skills work before deploymentStars: 323kUpdated Claude CodeCodexVerifiedInstall
xlsxanthropics/skillsUse this skill any time a spreadsheet file is the primary input or output. This means any task where the user wants to: open, read, edit, or fix an existing .xlsx, .xlsm, .xltx, .csv, or .tsv file (e.g., adding columns, computing formulas, formatting, charting, cleaning messy data); create a new spreadsheet from scratch or from other data sources; or convert between tabular file formats. Trigger especially when the user references a spreadsheet file by name or path — even casually (like "the xlsx in my downloads") — and wants something done to it or produced from it. Also trigger for cleaning or restructuring messy tabular data files (malformed rows, misplaced headers, junk data) into proper spreadsheets. The deliverable must be a spreadsheet file. Do NOT trigger when the primary deliverable is a Word document, HTML report, standalone Python script, database pipeline, or Google Sheets API integration, even if tabular data is involved.Stars: 198kUpdated Claude CodeCodexVerifiedInstall
shipping-and-launchaddyosmani/agent-skillsPrepares production launches. Use when preparing to deploy to production, or when asking what needs to be in place before shipping. Use when you need a pre-launch checklist, when setting up monitoring, when planning a staged rollout, or when you need a rollback strategy.Stars: 119kUpdated Claude CodeCodexVerifiedInstall
childhoodthedotmack/claude-memBabysit a shipped release after merge, publication, or deployment. Verify published packages, production behavior, installed plugins, and late CI or user reports until the observation window is clean; use babysit for PR readiness before release.Stars: 118kUpdated Claude CodeCodexVerifiedInstall