dev-skills

One iteration of hypothesis-driven bug discovery — audit the codebase against project invariants and risk patterns via parallel agents, adversarially verify each candidate, and file a groomed bug-issue (red-test sketch + verification evidence + proposed root-cause fix) for every finding that clears the gate. Does NOT implement or ship — specialized dev processes pick up the filed issues. Use when invoked as `/bughunt` (runs one iteration); use `/loop bughunt` to run it as an autonomous loop. Project-agnostic; reads risk dimensions from CLAUDE.md key-invariants and project memory.

Reference document used by the build-iteration skill /shovel-ready for its single-iteration machinery — branch, sync-with-master, /ship handoff, /retro pass, merge, and declaring the loop outcome. NOT user-invokable, and NOT the loop. The loop lives in `/loop`; this is one iteration's build cycle. The find-and-file siblings (/kaizen, /bughunt, /kaikaku) use `file-cycle` instead. Only load when a build-iteration skill directs you here.

Reference document shared by the find-and-file iteration skills (/kaizen, /bughunt, /kaikaku) for their common single-iteration machinery — author groomed issue(s), label by validation, file autonomously (multi-output), and declare the loop outcome. NOT user-invokable, and NOT the loop. The loop lives in `/loop`; this is one iteration's file cycle. The file-type sibling of `build-cycle`. Only load when a find-and-file skill directs you here.

Continuous architectural simplification and correctness-by-construction — whole-system recon via parallel agents to find high-leverage opportunities to simplify-and-beautify a structure OR foreclose a whole defect-class by making an illegal state unrepresentable, at the module/layer/concept altitude; steelman each candidate against Chesterton's Fence, then file groomed design-issue(s) for every candidate that clears the gate — autonomously, labeled by validation (no build, no in-session approval gate). Use when invoked as `/kaikaku` (runs one iteration); use `/loop kaikaku` to run it as an autonomous loop. The architectural-altitude sibling of `/kaizen`. Project-agnostic; reads simplicity + beauty + correctness stance from CLAUDE.md and project memory.

One iteration of accidental-complexity reduction — audit the codebase against craftsmanship principles via parallel agents, and file a groomed cleanup-issue for every candidate that clears the gate. Does NOT implement or ship — specialized dev processes pick up the filed issues. Use when invoked as `/kaizen` (runs one iteration); use `/loop kaizen` to run it as an autonomous loop. The instance/tidy-altitude sibling of `/kaikaku`. Project-agnostic; reads simplicity heuristics from CLAUDE.md and project memory.

Drive any single-iteration development skill as an autonomous loop — invoke one iteration, read its declared outcome, then pace, escalate, or stop. Use when invoked as `/loop <skill> [flags]` (e.g. `/loop kaizen --auto`, `/loop bughunt`, `/loop kaikaku`, `/loop shovel-ready --auto`). Owns cadence, idle escalation, and re-arming so the iteration skills stay pure. Pass `--once` to run exactly one iteration.

Reflect on a slice of the current session (a single iteration, the whole session, or a natural pause point) to identify durable, codifiable learnings about workflow OR dev infrastructure — and ship them only if they clear a quality bar. Most invocations produce nothing; that's the point. Use when the user asks to "run a retro", "do a retrospective", "reflect on the session", "what did you learn", "improve skills based on this session", "codify learnings"; when invoked from a loop driver (`/shovel-ready`) with `--scope=iteration`; or **proactively at pause points** when the agent is waiting on async work (CI, autodev runs, monitor events) and has cycles to think meta.

Drive a code change from working tree to a green PR — runs project checks, /simplify, code-review subagent, commits, opens PR, monitors CI. Stops when CI is green; caller handles merge. Use when invoked as `/ship`, or when `/shovel-ready` hands off a working tree of changes to be shipped. Project-agnostic; reads check commands from CLAUDE.md.

One iteration against a GitHub `shovel-ready` issue queue — if non-empty, take the highest-ROI issue (TDD, /ship, /retro, merge); if empty, audit closed work and unlabeled candidates to refill it; declare the outcome. Use when the user invokes `/shovel-ready`, asks to "work the shovel-ready queue", "drain shovel-ready", "drive issues to merge", "autopilot on labeled issues"; use `/loop shovel-ready --auto --drain` to clear the queue autonomously. GitHub-specific (uses `gh` CLI); skip for GitLab/Jira.

This skill should be used when the user asks to "commit", "push", "create a PR", "open a pull request", "monitor CI", "fix CI", "fix the build", "request review", "merge", "merge the PR", or mentions any git-based development workflow operations.

This skill should be used when the user asks to "diagnose", "debug", "investigate", "why is this broken", "find the root cause", "figure out why", "look at the logs", "what's going on", or when the user corrects with "you failed to get to the bottom of it" or questions apparent success. Also use when multiple attempts at fixing the same issue haven't worked, or when a problem needs systematic investigation.

This skill should be used when the user asks to "test this feature", "verify this works", "run tests", "check if this is working", "test thoroughly", or mentions testing a specific feature or fix.