V-Songbird
UserA custom marketplace of Claude Code plugins for both experienced developers and vibe coders, extending Claude Code with workflows, commands, and integrations that fit how you actually build.
Categories
Indexed Skills (19)
build-and-report
Aggregates implementer reports (or in-session edits if `/forge:dispatch-implementation` did not run), merges work-unit worktrees if any exist, bumps the version per the consuming project's convention, runs the project's build / verification commands, and produces a user-facing final report including the two mandatory sections "How to test this feature" and "How is this feature useful?". Stack-agnostic — actual build commands and version-bump file (`plugin.xml`, `package.json`, `Cargo.toml`, etc.) live in the consuming project's own CLAUDE.md.
critic-review
Dispatches the `adversarial-critic` subagent (Fable, read-only, bounded maxTurns) against the master plan in conversation context. The critic ground-truths the plan against the actual codebase and returns a structured critique (Blocking / High-priority / Open questions / Findings I couldn't ground in code / Verdict rationale). In deep mode, wraps the critic in a Workflow that adds a two-refuter panel per Blocking finding.
expert-analysis
Dispatches `forge-expert` subagents in parallel — one per chosen domain — to produce focused analyses of a feature against the codebase before a plan is drafted. Each expert covers one domain (architecture, performance, data/state, UI/UX, security, testing, build/tooling — pick from the role catalog) and returns a structured report citing `file:line` evidence.
forge
Creates new Claude Code rules and scaffolds .claude/rules/ files for specific topics or from project patterns. Writes new rule files only — never audits or scores existing rules (that's the assay skill), never modifies CLAUDE.md directly, and never writes code or tests.
plan-revise
Verifies each critique finding from `/forge:critic-review`, then folds the verified ones back into the master plan in conversation context. Walks the critic's blocking issues, high-priority gaps, and open questions; reads the cited code to confirm or refute each one; rewrites the affected plan sections; produces the final plan ready for user approval.
convention
Apply the Iceberg Convention — an asymmetric complexity management framework that quarantines system complexity below an architectural airgap, exposing a frictionless, hard-to-misuse surface above it for juniors and AI agents. Language- and framework-agnostic; Claude applies the rules as patterns and picks the idiomatic enforcement mechanism (types, linters, ADRs, FSMs, tracing) for whatever ecosystem the target code lives in.
router
Route Claude Code file-ops tools through a JetBrains IDE MCP server (WebStorm, Rider, IntelliJ IDEA). IDE reads reflect the editor's in-memory buffer so unsaved edits are visible; searches are scoped by the project index (skip node_modules, build outputs, .gitignore'd paths); and get_file_problems returns IDE inspection results without a cold language-server start.
init
Initializes kairoi on any project — detects tech stack, discovers modules, bootstraps model files from source scan. Supports Node (vitest/jest), JVM (gradle), Rust (cargo), Go, Python (pytest), and generic stacks.
kairoi
Code self-model methodology. Guards inject mechanically when I edit guarded files. Every commit is captured into buffer.jsonl automatically via the auto-buffer hook; reflection/sync is hook-dispatched, not user-invoked. Auto-activates in projects with .kairoi/model/_index.json.
lint
Observation-only report on source patterns that make Claude's own re-reading harder — star imports, files over 300 lines, source files with no matching test. Grounded in Claude's introspective knowledge of its own cognitive cost, not style-guide consensus.
assay
Rules quality auditor for existing Claude Code instruction files (CLAUDE.md, .claude/rules/). Grades, scores, audits, reviews, and critiques the structural clarity of existing rules. Reads and evaluates only — never creates new rules (that's forge), never reviews code or PRs, and ignores non-Claude-Code config like eslint or prettier.
file
Format and structurally clean Claude Code rule files. Applies four transforms: split multi-concept bullets, add blank-line separation, restructure long single-directive rules into sub-lists, and unwrap space-indented continuation lines. Never rewrites rule content, adds examples, or scores rules (those belong to assay and forge). Produces a before/after diff for user confirmation before writing.
primer
Installs the curated Claude Code rules primer into .claude/rules/recommendation-files.md. The bundled rules cover instruction-file hygiene: re-read CLAUDE.md, README.md, AGENTS.md, and .claude/rules/*.md after structural changes (renames, moves, deletions); verify file paths before citing them; and audit recommendation files for stale references when editing them. Copies one pre-defined file into .claude/rules/ — never authors new rules from scratch (use forge), audits existing rules (use assay), or reformats them (use file). When the destination already exists, the user is asked via AskUserQuestion buttons whether to overwrite, merge, or cancel.
audit
Reconciles kairoi model files against source code. Verifies guards, validates edges, resets confidence.
doctor
Validates structural integrity of .kairoi/ state files. Checks JSON validity, module consistency, guard fields, edge references, JSONL format, test command, and .gitignore coverage.
show
Pretty-print what kairoi is currently tracking about the codebase — modules, guards, edges, recent activity, pending overrides, buffer status. For the human auditing what Claude believes.
patterns
Senior-authored code patterns applied during code-writing, code-modifying, and refactoring tasks — branded/nominal types for domain identifiers, discriminated unions over optional-field records, finite state machines over boolean tuples, pure functions separated from I/O, explicit validation boundaries, exhaustive matching, sync-looking APIs, named call sites, no silent catches, no raw console logs. Adapts to surrounding codebase conventions first, then leans toward these patterns within the scope of the current change. Produces inline code only — does not generate ADRs, PR templates, lint configs, or audit reports.
scribe
Authors Claude Code instruction artifacts (SKILL.md, CLAUDE.md, plan files, subagent definitions, slash commands, hook scripts) with complete tool-call coverage — AskUserQuestion with header/multiSelect, TodoWrite/TaskCreate with paired content/activeForm, Bash with description, Agent with description/name, plan-mode gates, subagent limitations, hooks, permission modes — rather than prose placeholders or bare tool calls that Claude Code silently degrades.
status
Probe whether a JetBrains IDE MCP server is connected to this session and report the project modules it sees.
Bio shown is the top-scored skill's repo description as a fallback — real GitHub bios land in a future update.