p-wegner
@p-wegner
Public Skills
dependency-analyzer
by p-wegner
Analyze a ticket and its relationships to other open tickets, suggest dependency updates
endpoint-docs
by p-wegner
Create, maintain, and query a concise Markdown catalog of the REST API (docs/api/endpoints.md) — method, path, request shape, response, semantic — with a frontmatter SHA so staleness is detected via git diff. Use when asked "what endpoints exist", "is there an API for X", "where is endpoint Y used", "document the API", or after adding/changing Hono routes.
adversarial-arch-review
by p-wegner
Adversarial, code-metrics-driven architecture review. Distrust the codebase's own self-assessment (memory, "DONE/COMPLETE" notes, comments), use churn/complexity/coupling metrics to TARGET where to dig and to INFER structural problems, verify the most damning findings directly before relaying, and answer three questions — what architecture decisions truly limit future work, what dependencies (internal + external) are risky, what code is concerning. Use for "adversarial architecture review", "be critical about the architecture", "what's truly limiting us", "metrics-based code health audit", "where are the real risks".
quality-metrics-collector
by p-wegner
Collect repository code-health metrics and POST them to the board Quality Metrics API
agentic-kanban
by p-wegner
Kanban board for managing AI-driven coding tasks — cleanroom reimplementation of vibe-kanban
convert-hooks-to-pi
by p-wegner
Convert Claude Code hooks (.claude/settings.json hooks block + .claude/hooks/ scripts) into an equivalent Pi extension under .pi/plugin/. Use when porting a repo's hook automation from Claude Code to Pi.
kanban-workflow
by p-wegner
Guide for using the agentic-kanban MCP tools to reflect implementation progress on the board. Use when starting work on an issue, moving through implementation steps, reviewing completed work, or closing out a task.
e2e-author
by p-wegner
Scaffold a Playwright E2E test for agentic-kanban that follows the project's anti-flake rules from day one
scope-guard
by p-wegner
Pre-commit check for scope creep — diff working changes vs the task, flag unrelated files, run the project's fast checks, then commit. Stack-agnostic (reads the project's stack profile, not hard-coded to the TS monorepo).
merge-reconciler
by p-wegner
Land a whole BATCH of stranded/mergeable-but-conflicting workspaces onto the base branch in the most efficient way. You are handed the full batch as injected JSON ({{strandedBatch}}), analyze the whole set first (clean-independent / file-overlap clusters / migration collisions / dependencies), then land everything via the board's SAFE primitives — never manual git in the main checkout. Resolve each overlapping cluster's union ONCE in an integration worktree instead of N re-conflicting rebases. Report what landed and what was escalated.
fleet-analysis
by p-wegner
Time-scoped, fleet-level analysis of MANY agent sessions (Claude/Codex/Copilot) to find COMPOUNDING-ENGINEERING improvements — skills, hooks, helper scripts, deterministic board changes, CLAUDE.md edits. Use for "analyze the last 48h of sessions", "which tools fail most", "what are agents wasting tokens on", "where can we compound". Distinct from session-inspector (one session) and learning-step (one session → one fix).
ui-explorer
by p-wegner
"Visual UI exploration: find feature gaps, create tickets, fix stale docs about the featureset. Sources of truth are CLAUDE.md (operational) and docs/prd/01-features-catalog.md (catalog)."
java-backend-planner
by p-wegner
Explore codebase in parallel with subagents, interview user on requirements and constraints, then synthesize a phased Java backend implementation plan with tech stack decisions, module structure, risks, and DevOps strategy
db-doctor
by p-wegner
Diagnose & repair kanban.db migration/lock/WAL issues in place — never deletes, truncates, or resets the DB
convert-hooks-to-opencode
by p-wegner
Convert Claude Code hooks (.claude/settings.json hooks block + .claude/hooks/ scripts) into an equivalent OpenCode plugin under .opencode/plugin/. Use when porting a repo's hook automation from Claude Code to OpenCode.
flaky-test-triage
by p-wegner
Classify a failing test as known-flaky vs real regression, then re-run with the right strategy instead of debugging flakes or wrongly dismissing real failures
architecture-review
by p-wegner
Exhaustive architecture review — spawns parallel analysis agents, synthesizes findings, and creates kanban tickets for the top weaknesses
ui-review
by p-wegner
UI/UX review — spawns parallel agents to explore the running UI, identifies styling, workflow, and usability improvements, and creates kanban tickets
playwright-cli
by p-wegner
Automate browser interactions, test web pages and work with Playwright tests.
monitor-nudge
by p-wegner
Message sent to agents that have been running for more than 5 minutes without exiting — customize to change nudge behavior
architecture-improvement
by p-wegner
Analyzes the codebase using parallel subagents to identify high-value architecture improvement opportunities and produce a prioritized action plan.
orchestrator
by p-wegner
Delegating orchestrator — breaks work into sub-tasks and delegates to subagents or board tickets instead of doing everything itself
ticket-enhancer
by p-wegner
Enhance a ticket's title and description for clarity and completeness
--help
by p-wegner
--help