Software Development Skills

Use when the user invokes /cleanup-branch to repackage the current branch's commits into fewer, logically grouped, reviewable commits without changing the resulting code. Creates a backup branch first, verifies the regrouped tree is byte-for-byte identical to the original, prints the new commit list, and offers to delete the backup. Local only; never pushes. User-invoked only.

Use when the user invokes /fast-forward to land a finished feature branch (or its worktree) onto the local main/master branch as a fast-forward of its logically grouped, changelog-worthy commits. Commits any outstanding work, syncs and rebases onto the local trunk, regroups the history into a few reviewable commits whose messages read as public changelog entries, fast-forwards them onto the trunk with no merge commit, then deletes the branch and removes its worktree. User-invoked only; local only, never pushes; irreversible and destructive.

Build Python web applications with Flask 3+, using the application factory pattern and Blueprints. Use when developing Flask projects — blueprint structure, routes, authentication, error handling, working with sessions, implementing forms, configuring flask extensions, Jinja2 templates, CLI commands, logging, security, and deployment. Also use when the user mentions Flask routes, blueprints, app factory, Flask-Login, Flask-WTF, or any Flask extension, even if they don't explicitly say "Flask development.

Use when writing, debugging, or reviewing htmx attributes (hx-get, hx-post, hx-swap, hx-target, hx-trigger, hx-boost), AJAX interactions, server-side HTML fragment responses, out-of-band swaps, htmx events, loading indicators, or hypermedia-driven patterns. Also use when the user mentions htmx class names like htmx-indicator, htmx-request, or any hx- attributes, even if they don't explicitly say "htmx.

Use whenever writing or editing Python code in a project that depends on the nclutils package. Trigger on any import nclutils or from nclutils.

Use when the user invokes /prune-comments to clean up inline code comments. With no arguments it reviews the current changes (uncommitted working-tree changes, or the branch-vs-trunk diff when the tree is clean); with arguments it reviews the named files, directories, or globs in full. Deletes redundant what-comments, tightens verbose ones, and keeps genuine why-comments, leaving noqa/type:ignore and other tooling directives untouched. Edits the working tree in place and does not commit. User-invoked only.

Triage and curate this project's persisted recall backlog - validate the open items against the current repo, apply the resulting fixes (close finished work, remove obsolete items, correct drifted ones), then rank what remains by impact and effort to recommend what to do next. Pass --clean to apply removals automatically instead of confirming them.

Backfill this project's recall memory store from past Claude Code session transcripts, mining prior conversations for durable learnings and deferred work.

Write a consume-once handoff of the current session's working state into this project's recall store, so the next session picks up where this one left off. Most useful right before /compact or /clear. The next fresh session (compact, clear, or startup) injects the handoff and then deletes it.

Review and curate this project's persisted memory store - deduplicate learnings, prune stale entries, correct frontmatter, and close resolved backlog items. Pass --clean to apply the changes automatically instead of proposing them.

Build Python applications using Tortoise ORM v1.x, the async Python ORM inspired by Django. Use when writing or modifying Tortoise Model definitions, queries, relations (ForeignKeyField, ManyToManyField, OneToOneField), transactions, signals, validators, Pydantic integration, or Tortoise initialization. Also use when the user mentions Tortoise ORM, tortoise-orm, async ORM with asyncpg/asyncmy, fields.ForeignKeyField, fields.ManyToManyField, prefetchrelated, selectrelated, Tortoise.init, RegisterTortoise, tortoise.contrib.pydantic, or any Tortoise ORM patterns, even if they don't explicitly say 'Tortoise.' Covers the full lifecycle, model definition, CRUD, relations, querying, transactions, signals, migrations, testing, and Pydantic serialization.

Use when authoring, building, configuring, or reviewing documentation with the Zensical static-site engine. Trigger on any zensical.toml or mkdocs.yml work, running zensical new/build/serve, writing docs/ Markdown pages, or requests to make documentation richer with Mermaid diagrams, admonitions/call-outs, code annotations, content tabs, grids, or icons. Also use whenever writing Markdown that will be published with Zensical, even if the user does not say 'Zensical' by name. Actively reach for these rich elements whenever they make the docs clearer, not only when asked.