Development plugins

Architecture Decision Records: init, distill from existing code, record, review diffs, validate. Stack-agnostic.

5 skills

A capture-to-ship development loop for any git repository agents work in: loop documents, one config file, and checks and hooks that enforce the rules.

1 skill · 3 commands · hooks

Starter specialist agents: architect, backend, frontend, design and devops.

5 agents

Multi-agent iterative code review until zero issues remain

1 skill · 3 commands

Configure and report bounded, repository-specific coding-agent signals to Pomegr.

2 skills · 1 MCP server · hooks

One skill carrying the whole discipline, two agents, per-host session protection. Plan first, build clean, keep every change on the record. stem.cooking

1 skill · 2 agents · hooks

Browser automation MCP + Continuum context-chain memory + in-page hint modal. Pair with the Mochi Chrome extension (⌘⇧M on any tab).

1 skill · 3 MCP servers

Get your agent's real runs into AgentX, then score them, then turn what they measure into a code fix. /instrument sets up a Python agent and proves it with traces it reads back; /run-eval evaluates it against a dataset it can also create - templates, files, or cases curated from live traces; /eval-fix triages the evaluation against the source and re-runs it on the same dataset.

4 skills

Standalone multi-agent orchestration via pipeline.yml (fork coders + reviewer/validator loops).

1 skill · 3 agents

AI-assisted deep planning with research, interview, external LLM review, and TDD approach

1 skill · 2 agents · hooks

Up-to-date library documentation lookup via Upstash Context7 MCP server

4 skills

Build agent-step LangGraph agents — a ReAct agent fronted by one batching tool over a deterministic flow-controller runner (vendored and versioned; agent-step 3.0.1). Bootstrap a project (runner library + streaming CLI + graph/state/prompt scaffold + the required local sandbox service), then add domain tools declaratively: each action declares its params, prereqs and verdict ROWS — summary, doctrine, state write, effects, one authority per outcome — and its executor just names the verdict. The engine owns the machinery and the model-facing language it enforces: composed action descriptions, per-gate reply contracts, the spliceable turn-protocol prompt fragment, and engine-rendered read-backs and standing asks the model speaks verbatim. Mechanisms are configuration, not prompt prose — confirmation propose/execute gates with attempt budgets and stored-byte repetition, OTP and double-entry match gates, multi-turn flows, paginated reads, escalation ladders that count the prompt's "once" rules and escalate themselves, capture-bounce bounds, an engine-enforced abort policy, typed handoff routes, and a state-resolved handback envelope with orchestrator/specialized agent roles plus off-topic delegation. Graph state is one Zod schema (reducers + invoke-boundary validation, channel-completeness enforced at construction). Also: caller-capture param primitives, the twelve measured host-architecture design principles, porting an existing project onto agent-step, Kafka run-tracing observability via the companion kafka-observability plugin, the three-layer test methodology (runner unit / sandbox / prompt-input), and a channel-middleware contract audit.

4 skills

A mini-foundry — the build-verify-fix engine without the MCP server or the planning interview. You bring a structured spec (markdown with stable requirement IDs + acceptance criteria); crucible builds it and refuses to let it drift. /crucible:build <spec-file> decomposes the spec into units with DISJOINT file ownership, shows the split for one approval, then runs a Workflow engine: CAST (parallel deliberate-engineering builders, each handed its unit prompt verbatim, each emitting re-executable evidence bound to requirement IDs) → INSPECT (three blind orthogonal streams — wiring exists/substantive/wired, spec-before-code fresh-eyes assay, and re-execution of every evidence command) → GRIND (every non-passing verdict is a defect; fix the affected units and full-re-inspect, loop until zero or a cycle cap) → ASSAY (fresh-eyes agents read the spec first, form expectations, then confirm each requirement). Refuses a dirty tree; edits the shared working tree under disjoint ownership; commits per-unit only on user say-so so any unit is one revert away. Keeps foundry's load-bearing principles — verbatim unit prompts (no interpretation layer), multi-stream blind verification (build-green ≠ done), evidence binding, defect-to-zero grind, spec-before-code assay, an engine that routes and verifies but never re-narrates the spec or judges the code itself. Drops what needs the heavy machinery: no MCP gates/hashing, no isolated-worktree evidence, no brownfield flow-graph interview. forge plans → foundry builds; crucible is foundry's build engine, mini.

2 commands · 2 agents

Translate a validated AGER/OKF AgentGraph into a Grok Build plugin.

3 skills · 1 command

Translate a validated AGER/OKF AgentGraph into a Codex plugin.

3 skills · 1 command

One skill carrying the whole discipline, two agents, per-host session protection. Plan first, build clean, keep every change on the record. stem.cooking

1 skill · 2 agents · hooks

The physical AI platform for robot training, run from your coding agent: data collection and telemetry, real-to-sim system identification, simulation, synthetic data, reinforcement and imitation learning, evaluation, sim-to-real deployment, and the data flywheel back into training. One MCP server, the Studio (downloaded on first use), seven agents and two skills.

2 skills · 1 MCP server · hooks

Standalone documentation sync: find outdated refs, update CHANGELOG, flag stale examples based on code changes

1 skill · 1 command · 1 agent

Run any model cheaper than the advisor (Haiku, Sonnet, or Opus) as the executor and escalate hard decisions to a Fable 5 advisor. Near frontier-level intelligence at executor-level cost.

1 skill · 1 agent · hooks

Git recovery, branch handoffs, and bounded repo context in one plugin. Use /taf:work-recovery, /taf:branch-handoff, or /taf:prepare-repo-context.

3 skills

Implement code from /deep-plan sections with TDD, code review, and git workflow

1 skill · 1 agent · hooks

Studio Control Center plugin for Claude Code with integrated skills for code analysis, optimization, and best practices.

3 skills

Initialize and manage Nix flake development environments with auto-detection and security tooling

1 skill · 1 MCP server · hooks

Native Rust MCP with persistent Python, bounded output, observable cancellation, shell composition and an explicitly configured MCP broker. Bundles a usage skill and nudge hook.

1 skill · 1 MCP server · hooks

GitHub conventions -- branch naming, commit format, issue/PR templates, and safe issue/PR referencing

2 skills · hooks

Depth, surface and topology inspection of Blender scenes: mesh defect audits, depth-revealing renders and depth-ordering reports read from the scene itself.

2 skills · 1 agent

Control a local Chromium browser and inspect transient behavior with source-linked temporal visual evidence.

2 skills · 1 MCP server

Wax agent skills for registry search, source-backed implementation, registry authoring, and scan analytics

3 skills

Review iOS 27 engineering, design and store assets, audit submissions, recover from rejections, and evaluate Guideline 4.3 craft risks.

2 skills · 1 command

Generate, validate and simulate Formhaus form definitions, and bind a Figma design system to the Formhaus Figma plugin.

2 skills · 1 MCP server