Skills and workflows for working with @outfitter/* packages. Patterns, templates, compliance checking, and debugging for Outfitter.
Work with @outfitter/* packages — implement handlers, adopt patterns, review compliance, debug issues, and update versions. Routes to the right skill based on task type.
Debug @outfitter/* issues — Result handling, MCP, CLI, logging, exit codes. Evidence-based investigation, not trial-and-error.
Systematic debugging process for @outfitter/* package issues. Use when debugging Result handling, MCP problems, CLI output, exit codes, logging, or unexpected behavior. Produces structured investigation reports with root cause analysis.
Generates patterns, templates, and guides for @outfitter/* packages. Covers transport-agnostic handler systems, Result types, error taxonomy, and package APIs. Use when working with @outfitter/*, Result types, Handler contract, error taxonomy, or when Result, Handler, ValidationError, NotFoundError, OutfitterError, or package names like contracts, cli, mcp, schema, tui, daemon, config, logging are mentioned.
Verify Outfitter compliance in a codebase. Scans for anti-patterns (throws, console, hardcoded paths) and produces a severity-ranked compliance report. Use for pre-commit checks, code review, or migration validation.
Submit feedback to the Outfitter team via GitHub issues. Use after discovering bugs, missing features, unclear docs, or improvement opportunities in @outfitter/* packages.
Start with Outfitter — scaffold new projects or adopt patterns in existing codebases. Detects context, scans for adoption candidates, and orchestrates phased conversion.
Uses power tools
Uses Bash, Write, or Edit tools
Own this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimOwn this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimBased on adoption, maintenance, documentation, and repository signals. Not a security audit or endorsement.
Outfitter is shared infrastructure for building AI-agent-ready tools. CLIs, MCP servers, daemons—with patterns that work for both humans and machines.
Status: Active Development
bunx outfitter init my-project --preset cli --yes
cd my-project
bun install
bun run dev
Output:
my-project v0.1.0
Usage: my-project [command] [options]
Commands:
example Run example command
help Display help for command
That scaffolds a working CLI with typed commands, output mode detection, and proper error handling. Four commands to useful software.
We kept solving the same problems across projects: config loading, error handling, CLI output modes, MCP server boilerplate. Every tool needed the same foundation.
So we extracted it.
Outfitter is opinionated infrastructure for the Bun ecosystem. The patterns assume you're building tools that agents will consume—structured output, typed errors, predictable behavior. Humans benefit too; agents just make the stakes clearer.
The core idea: handlers are pure functions returning Result types. CLI and MCP are thin adapters over the same logic. Write the handler once, expose it everywhere.
@outfitter/cli, @outfitter/contracts, @outfitter/mcp) — NDJSON output with --stream, transport-agnostic ctx.progress on HandlerContext, and MCP progress adapter support (notifications/progress).@outfitter/cli, @outfitter/mcp) — .destructive(true) auto-adds --dry-run, .readOnly(true) / .idempotent(true) metadata maps to MCP annotations, and error envelopes include retryable + retry_after.@outfitter/cli) — output truncation adds pagination hints and full-output file pointers, and .relatedTo() builds an action graph for tier-4 related-command hints.examples/reference/) — end-to-end example showing v0.4-v0.6 patterns across shared handlers, CLI, and MCP.For migration details and before/after examples, see v0.6 Migration.
@outfitter/cli) — .input(schema) auto-derives flags from Zod, .context(factory) adds typed async context, .hints() / .onError() keep hints transport-local, createSchemaPreset() enables schema-driven presets, and runHandler() + output.envelope() unify lifecycle/output/exit mapping.@outfitter/cli) — self-documenting root command and tiered hint generation (command tree, error category recovery, schema-derived params).@outfitter/contracts) — canonical ActionHint / CLIHint / MCPHint types plus enriched ErrorCategory metadata via errorCategoryMeta().@outfitter/mcp) — defineResourceTemplate() now supports Zod-validated typed URI template params.loadConfig() schema argument is now optional; testCommand() / testTool() include new input, context, json, envelope, and hints support.outfitter check action-registry, outfitter upgrade --codemods, and richer starter examples (outfitter init cli ... --example todo, outfitter init mcp ... --example files).For migration details and before/after examples, see v0.5 Migration.
Three tiers, one goal: shared infrastructure that works for humans and machines.
APIs everything else depends on.
| Package | What it does |
|---|---|
@outfitter/contracts | Result/Error patterns, handler contract, and utilities like parseInput(), wrapError(), fromFetch(), expectOk(), expectErr() |
@outfitter/types | Type utilities and branded types |
The building blocks for applications.
npx claudepluginhub outfitter-dev/outfitter --plugin outfitterFast local documentation search with llms.txt indexing. Search 12K+ line docs in 6ms with line-accurate citations. One command (/blz) and one agent (@blz:blazer) for all documentation operations.
Internal Outfitter team skills for docs, voice, and conventions
Development methodology fieldguides for Claude Code. TDD, debugging, architecture, research, multi-agent coordination, plus authoring skills for plugins, agents, commands, and hooks.
Build with the Trails framework — contract-first trails, trailheads, testing, and governance for agent-assisted development.
Waymark annotation grammar and tooling for code-adjacent context
MCP server development helper with tool and resource scaffolding
This skill should be used when the model's ROLE_TYPE is orchestrator and needs to delegate tasks to specialist sub-agents. Provides scientific delegation framework ensuring world-building context (WHERE, WHAT, WHY) while preserving agent autonomy in implementation decisions (HOW). Use when planning task delegation, structuring sub-agent prompts, or coordinating multi-agent workflows.
Transform Claude Code into a structured development platform with 29 /sc: commands, 23 specialized agents, 7 behavioral modes, and MCP server integration
Unified capability management center for Skills, Agents, and Commands.
Harness-native ECC operator layer - 67 agents, 271 skills, 92 legacy command shims, reusable hooks, rules, selective install profiles, and production-ready workflows for Claude Code, Codex, OpenCode, Cursor, and related agent harnesses
v9.44.1 — Patch release for Gemini environment/version detection and qwen auth gating. Run /octo:setup.