Introduce codespell spell-checking to a project. Creates branch, config file, CI workflow (GitHub Actions or Forgejo Actions for Codeberg), pre-commit hook. Lists typos, helps identify paths/words to ignore, and fixes typos interactively. Use when setting up codespell in a new project.

2026-06-192

introduce-reuse-compliance

con/skills

Introduce REUSE specification compliance (LICENSES/ directory, REUSE.toml, SPDX headers) to a software project or BIDS dataset, then validate it. Covers BIDS data-vs-code separation, DUO (Data Use Ontology) integration, DEP-3 patch tagging for vendoring repos, and integration with tox / pre-commit / Makefile / GitHub Actions. Use when adding licensing metadata to a project, fixing `reuse lint` failures, licensing a BIDS dataset, or annotating patches in a vendoring repo.

2026-06-012

introduce-git-bug

con/skills

Introduce git-bug distributed issue tracking to a git project. Configures GitHub bridge, syncs issues, pushes refs, and documents workflow in DEVELOPMENT.md and CLAUDE.md. Use when setting up git-bug in a new project.

2026-05-172

credit-contributions

con/skills

Annotate a paper or work with CRediT (Contributor Roles Taxonomy) author contributions and render the LaTeX byline. Creates a `.tributors.credit.yaml` overlay (single source of truth, compatible with con/tributors), seeds it from sibling repos' git history when present, renders the "Author Contributions" section into LaTeX or Markdown for the target venue, emits JATS XML for publisher metadata, and (via render_authors.py) renders the `\author{}` block with affiliation references. Use when a paper needs a CRediT statement, when standardising contributor attribution across a multi-repo work, when authors need an auto-generated byline + affiliation block, or when the user mentions CRediT / NISO Z39.104 / contributor roles / author contribution section.

2026-05-162

pr-feedback-review

con/skills

Load a PR's review feedback (human + bot) **and CI status**, classify each comment, and recommend what to address vs dismiss with draft responses. CI failures (merge conflicts, lint, tests, workflow runs) are first-class — diagnose, propose a fix, and bundle into the same actionable report. Works from a local repo directory or a PR URL.

2026-05-082

pr-review-update

con/skills

Review dashboard PRs needing your response and generate high-confidence update proposals for PRs where maintainers are waiting

2026-04-292

name	scan-projects
description	Scan subdirectories and files to create/update projects.tsv with metadata and LLM-generated summaries

Scan Projects Skill

Scans all entries (git repos, plain directories, and standalone files) in the current folder and creates/updates a projects.tsv file with metadata and summaries.

Output Format

The projects.tsv file contains tab-separated columns:

folder: Entry name (directory or filename)
type: Entry type — git, dir, or file
summary: High-level description
language: Primary programming language (or file type for standalone files)
license: License type (git repos and directories only)
earliest_commit: Date of first commit (git repos only, ISO format)
latest_commit: Date of most recent commit on main/master (git repos only, ISO format)
url: Remote git URL (git repos only)

Execution Instructions

Phase 1: Scan Metadata

python3 ~/.claude/skills/scan-projects/scan.py

Collects metadata for all entries. Git repos get full metadata (language, license, commits, URL). Plain directories get language detection and license scanning. Files get type classification. All summaries start as "NEEDS_ANALYSIS".

Phase 2: Generate Summaries with Claude Analysis

Read projects.tsv to find entries with "NEEDS_ANALYSIS"
Batch process entries (20+ at a time using parallel Explore agents)
For each entry, analyze to determine purpose/goal:
- Git repos/dirs: Read README, main code, package metadata, directory structure
- Files: Read content (if text), infer purpose from name and context
Generate a concise summary (1-2 sentences, max 150 chars):
- Focus on purpose and goal, not implementation
- Be specific about the domain
- Avoid generic phrases — just state what it does
Batch update the TSV via inline Python or batch_update.py

Summary Guidelines

Good: "BIDS validator for neuroimaging datasets"
Good: "CLI tool for downloading CI logs from GitHub Actions, Travis, Appveyor"
Good: "GPG-encrypted passwords/credentials file"
Bad: "This is a Python library"
Bad: "A directory with some files"

Helper Scripts

update_summary.py --list [N]: List N entries needing analysis
update_summary.py <folder> <summary>: Update single entry
batch_update.py -: Batch update from JSON (stdin)
batch_update.py <file.json>: Batch update from JSON file

Entry Type Handling

Type	Language detection	License	Commits	URL
`git`	File extension counting	LICENSE/COPYING parsing	Earliest + latest	`git remote get-url origin`
`dir`	File extension counting	LICENSE/COPYING parsing	N/A	N/A
`file`	Extension mapping	N/A	N/A	N/A

Error Handling

Report entries where scanning fails
Handle missing LICENSE/README gracefully
Handle repositories with no commits
Skip hidden directories and .claude